Civil Engineering Contract Administration and Control, Second Edition I.H. Seeley
Highway Tratlic Analysis and Design, Third Edition R.J. Salter and N. Hounsell
Plastic Analysis of Steel and Concrete Structures, Second Edition S.J.J. Moy
Reinforced Concrete Design, Fourth Edition W.H. Mosley and J.H. Bungey
Reinforced Concrete Design to Eurocode 2 W.H. Mosley, R.Hulse and J.H. Bungey
A catalogue record for this book is available from the British Library.
This book is printed on paper suitable for recycling and made from
fully managed and sustained forest sources.
10 9 8 7 6 5 4 3 2
07 06 05 04 03 02 01 00 99 98
To my parents Jeannie and Robert
Contents
Preface Xll
Acknowledgements Xlll
1. Structural Steelwork 1
1.1 Introduction 1
1.2 The Design Process 5
1. 2.1 Aims of Design 5
1. 2. 2 Structural Loading 6
1. 2. 3 Structural Analysis and Load Distribution 8
1.3 Example 1.1 Load distribution oneway spanning slabs 8
1.4 Example 1.2 Twoway spanning slabs 9
1.5 Example 1.3 Secondary beams 11
1.6 Example 1.4 Combined oneway, twoway slabs and beams 12
1.7 Limit State Design BS 5950:Part 1: 1990 13
1. 7.1 Ultimate Limit State 14
1. 7.2 Serviceability Limit State 14
1. 7.3 Partial Safety Factors 14
1. 7. 4 Application ofPartial Safety Factors 14
1.8 Example 1.5 Brewery part floor plan load distribution 15
1.9 Application ofWind Loads 18
1.10 Example 1. 6 Storage hopper 18
1.11 Example 1. 7 Industrial warehouse 22
1.12 Example 1.8 Radar reflector 32
2. Flexural Members 34
2.1 Introduction 34
2.1.1 PlasticSections 37
2.1.2 Compact Sections 38
2.1.3 SemiCompact Sections 38
2.1.4 Slender Sections 38
2.2 Shear Capacity (Clause 4.2.3) 39
2.3 Example 2.1 Shear check of a simply supported beam 40
2.4 Moment Capacity (Clause 4.2.5) 41
2. 4.1 Compression Flange Restraint 41
2.4.2 Effective Length L. (Clauses 4.3.5 and 4.3.6) 44
2. 4. 3 Moment Capacity (Me) ofBeams with Full Lateral Restraint 46
2.5 Exan1ple 2.2 Bending in fully restrained beam 47
2.6 Moment Capacity (Mb) of Beams without Full Lateral Restraint
(Clause 4.3. 7) 48
2.6.1 Rigorous Method (Clause 4.3. 7.1) 48
2.7 Example 2.3 Beam with intermittent lateral restraint 49
Vll
Vlll Contents
2.8 Example 2.4 Beam with intermittent lateral restraint 53
2.9 Example 2.5 Rectangular hollow section as a beam 56
2.10 Example 2.6 Cantilever beam 57
2.11 Web Buckling and Web Bearing 60
2.11.1 Web Buckling (Clause 4.5.2) 61
2.11.2 Web Bearing (Clause 4.5.3) 62
2.12 Example 2.7 Web bearing and web buckling at support 63
2.13 Deflection ofBeams (Clause 2.5.1) 65
2.1 3.1 Equivalent UDL Technique 66
2.14 Example 2.8 Deflection of simply supported beam 68
2.15 Conservative Method for Lateral Torsional Buckling Moment
Capacity (Clause 4.3. 7. 7) 70
2.16 Example 2.9 Simply supported beam 1 conservative method 70
2.17 Example 2.10 Simply beam 2  conservative method 71
2.18 Safe Load Tables 72
2.19 Example 2.11 Beam with intermittent lateral restraint  Use of safe
load tables 73
2.19.1 Shear 75
2.19.2 Bending 75
2.19. 3 Web Buckling 75
2.19.4 Web Bearing 76
2.19.5 Deflection 76
2.20 Example 2.12 Pedestrian walkway 77
2.21 Solution to Example 2.2 78
2.22 Solution to Example 2.12 82
5. Connections 145
5 .1 Introduction 145
5.1.1 Simple Connections 146
5.1.2 Moment Connections 147
5. 2 Bolted Cmmections 153
5. 2.1 Black Bolts 153
5.2.2 High Strength Friction Grip Bolts (H.S.FG.) 156
5. 2. 3 Design of Simple Connections 158
5. 3 Welded Connections 159
5. 3.1 Fillet Welds· 160
5.3.2 Butt Welds 161
5. 3. 3 Design ofFillet Weld Connections 162
5 .4 Beam End Connections 164
5. 4.1 DoubleAngle Web Cleats 164
5. 4. 2 Flexible End Plates 165
5.4.3 Fin Plates 165
5.5 Example 5.5 Web cleat, end plate and fin plate connections 166
5.6 Design of Moment Connections 167
5. 6.1 Typical Site Connection Using H. SF G. Bolts 168
5. 6. 2 Example 5. 6 Moment connection in rectangular portal frame 168
5. 6. 3 Example 5. 7 Crane bracket moment connection 170
5.7 Splices 171
5. 7.1 Beam Splices 171
5. 7. 2 Example 5. 8 Beam splice 172
5. 7. 3 Column Splices 178
5. 7. 4 Example 5. 9 Column splice 180
5. 8 Solution to Example 5.1 183
5.9 Solution to Exan1ple 5.2 185
5.10 Solution to Example 5.5 187
5. 11 Solution to Example 5. 6 196
X Contents
6. Plate Girders 198
6. 1 Introduction 198
6.1.1 Design Load Effects 200
6.2 Initial Sizing 201
6.3 Moment and Shear Capacity 201
6 .4 Deflection 203
6. 5 Intermediate Stiffeners 203
6. 6 Load Bearing Stiffeners 203
6.7 Example 6.1 Plate girder in multistorey office block 203
6. 7.1 Design Loading 205
6. 7. 2 Column Loads 205
6. 7. 3 Initial Sizing 206
6. 7.4 Section Classification (Clause 3.5) 207
6. 7. 5 Flanges 207
6. 7.6 Web 208
6. 7. 7 Moment Capacity (Clause 4.4.4.2(a)) 209
6. 7.8 Shear Capacity (Clause 4.4.5.3) 209
6. 7.9 Deflection (Clause 2.5) 210
6. 7.10 Intermediate St~ffeners (Clause 4. 4. 6) 212
6. 7.11 Load Bearing St~!Jeners (Clause 4. 5) 213
6. 7.12 Flange to Web Connection 217
6.8 Example 6.2 Plate girder with intermittent restraint to flange 218
6.9 Solution to Example 6.2 219
Bibliography 270
Index 272
Preface
W.M.C. McKenzie
Xll
Acknowledgements
The author and the publishers are grateful to the following organisations for permission to
reproduce copyright material in this book.
Extracts from BS 5950: Part 1: 1990 are reproduced with the permission of BSI. Complete
editions of the standards can be obtained by post from BSI Customer Services, 389
Chiswick High Road, London W4 4AL.
Xlll
1. Structural Steelwork
1.1 Introduction
Modern steelmaking processes employing blast furnaces to produce pigiron from ironore,
coke and limestone are refmements of those methods used by early ironworkers. Despite
smelting techniques being used for several thousand years it is only relatively recently (i.e.
the 1800s), with William Kelly's 'pneumatic' steelmaking process in America, and shortly
after, Sir Henry Bessemer's converter in England, and the SiemensMartin openhearth
process, that methods of mass production of steel have evolved.
Essentially, the raw materials are heated to temperatures in the region of 1500°C. During
this heating process the coke releases carbon monoxide, which when combined with the iron
oxides in the ore, produces metallic iron. Impurities (e.g. silica which combines with the
limestone to form calcium silicate) form a slag which floats to the top of the molten metal
and is removed leaving 'pigiron' behind. Further refinements of the pigiron, in which it is
combined with scrap steel and iron ore are carried out to remove excess carbon and other
impurities such as silica, phosphorus, manganese and sulphur. The resulting molten steel
can be poured into castiron moulds to form ingots or continuously cast into slabs which are
subsequently cut to required lengths.
The precise method of refinement used depends on the type of steel required. There are
five main categories:
(i) carbon steels containing varying amounts of carbon, manganese, silica and copper;
(ii) alloy steels as in (i) and in addition containing varying amounts of vanadium,
molybdenum or other elements;
(iii) highstrength lowalloy steels as in (i) and in addition containing small amounts of
expensive alloying elements;
(iv) stainless steels as in (i) and in addition containing chromium, nickel and other
alloying elements;
(v) tool steels as in (i) and in addition containing tungsten, molybdenum and other
alloying elements.
Each type of steel exhibits different characteristics such as strength, ductility, hardness,
and corrosion resistance. The most widely used type of structural steelwork is carbon steel,
although stainless steel is frequently used for architectural reasons and has been used on
some of the world's most famous structures, e.g. the Chrysler Building in New York and the
Canary Wharf Tower in London. In addition to aesthetic considerations, stainless steel
significantly reduces the maintenance requirements and costs during the lifetime of a
structure.
The physical properties of various types of steel and steel alloys depend primarily on the
percentage and distribution of carbon in the iron; in particular the relative proportions of
ferrite, pearlite and cementite, all of which contain carbon. Heat treatment of steel is carried
out to control the amount, size, shape and distribution of these substances and to develop
2 Design ofStructural Steelwork
other compounds such as austenite and martensite. This treatment hardens and strengthens
the steel but in addition induces additional residual strains. Following heat treatment,
tempering (annealing) is normally carried out to relieve these strains. This consists of
reheating to a lower temperature which results in a reduction in hardening and strength but
produces an increase in ductility and toughness.
The first major structural use of modem steel in Britain was in the construction of the
Forth Railway Bridge in Scotland. This impressive elegant structure designed by Baker and
Fowler makes use of tubular and angular sections riveted and forged together to form the
appropriate structural elements. The development of hotrolling, welding and bolting
techniques during the twentieth century has resulted in a vast range of available cross
section and fabrication options which enable designers to produce a limitless variety of
structures reflecting the lightness, stiffuess and strength of modem steel.
Steel is in many respects an ideal structural material which exhibits a number of useful
characteristics such as:
In addition, the plastic deformation which steel structures exhibit after initial yield gives
a margin of safety and early warning that the structure is stressed beyond the elastic limit,
and that loads should be removed and the structure should be strengthened.
As a result of these properties, structures in steelwork can be analysed using a variety of
classic elastic analysis techniques, elastoplastic and plastic methods. The deformations and
stresses predicted using such techniques have good correlation with experimentally
measured values.
There are some aspects of the behaviour of steel structures which require more rigorous
analysis than is possible using the principles of statics and linear elastic techniques. When
structural steel sections, which may be hotrolled or cold formed into thin sections, are
subject to stress concentrations at locations such as support points or reentrant comers,
local yielding may occur with a subsequent redistribution of loading through the structure.
A result of this is that these structural elements are susceptible to failure by buckling under
compression, lateral torsional buckling (bowing laterally along the member length) or
bearing (crinkling) locally under concentrated loads.
Steel is a material which offers a designer wide scope to create original and unique
structural forms and systems. There are, however, several structural forms which have
evolved to fulfil clearly defined requirements, and which have a widely adopted range of
standardised structural sections and connection details. In addition to the requirements of
economy and speed of construction on site, a client may have specific reasons for choosing
a steelframed structural system; for example a large columnfree internal space, ease in
Structural Steelwork 3
+ Steelrolling mills which are highly automated, as are the fabrication workshops,
thereby improving product quality and reducing the cost of finished steel
sections for delivery to site,
+ Specialist companies exist which design, fabricate and in some cases, deliver
and erect, widely used special structural members and systems,
+ A variety of special alloy steels for specific functional or aesthetic applications
have been developed. Many structural steel sections are for example available in
several grades of stainless steel containing a percentage of chromium and/or
nickel for use in hostile corrosive environments. Weathering steel contains a
small percentage of copper which permits an outer layer of corrosion to develop
on the exposed surfaces which inhibits further deterioration within the section,
+ Fully automated, computercontrolled cutting and welding of steel sections,
particularly rolled hollow sections, has led to the development of prefabricated
trusses and spaceframe modules. The extensive use of HighStrength Friction
Grip bolts and to a lesser extent site welding, is now widely adopted to achieve
efficient and economic rigid connections.
The two most widely used applications of structural steelwork in buildings are:
(ii) Multistorey braced frames in which the lateral wind loads are resisted by steel
bracing elements or reinforced concrete shear walls as shown in Figure 1.2, or
alternatively rigidjointed frames in which lateral stability is provided by moment
connections between the columns and floor beams.
Both systems have advantages and disadvantages; the most appropriate one for any
given structure will depend on a number of factors, such as the need for adaptability, speed
and ease of fabrication, contract programme constraints and fireproofing requirements.
Figure 1.1
Figure 1.2
Structural Steelwork 5
Some types of building have a structural form and proportions which lend themselves to
a particular system, but there is inevitably some overlap between structural systems and
applications.
The adoption of a particular structural form or system may be the result of economic or
architectural style as much as structural and functional necessity. In the 1970s significant
investment in modernising and renewing urban light rail systems resulted in a number of
bright, attractive rail stations with space deck roofs, such as those built for the Newcastle
Metro and the Glasgow Underground railways. In the 1980s there was a major drive to
develop London's Docklands as a new commercial and financial centre, resulting in the
widespread construction of 'fasttrack' multistorey steel frames with composite floor
construction. In the 1990s, extensive reconstruction of football stadia has taken place
following the Taylor Report (Ref:21) on crowd safety at football grounds. Again steel
framed construction has been adopted to take advantage of the speed of erection.
The design of a structure is initiated by the acceptance of a client's brief defining specific
requirements. The brief will generally indicate the use which is to be made of the
structure(s), e.g. an office block with specified usable floor areas, a commercial
development utilizing a limited plot of land, or a transportation development scheme
comprising a variety of structures such as bridges, terminal buildings and associated
facilities.
Preliminary designs and scheme drawings are prepared and considered by the design
team, which may include structural engineers, architects, local authority representives, the
potential user and the client.
After evaluation of the possible alternatives, including an estimate of budget costs and
approval by the client, final detailed design and preparation of working drawings is carried
out. Most contracts go through a tender stage in which a number of contractors are asked to
submit a tender price (bid) to carry out the construction work; this sometimes involves
design and build arrangements. There are numerous variations in the precise details for
which bids are made. Site supervision of the successful contractor is normally carried out
by a representative of the consultant engaged by the client.
All structures are subjected to loading from various sources. The main categories of loading
are: dead, imposed and wind loads. In some circumstances there may be other loading types
which should be considered such as settlement, fatigue, temperature effects, dynamic
loading, impact effects, e.g. when designing bridge decks, cranegantry girders or maritime
structures. In the majority of cases design considering combinations of dead, imposed and
wind loads is the most appropriate.
The defmition of dead and imposed loading is given in BS 6399:Part 1:1984, while wind
loading is defined in BS 6399:Part 2:1995. Part 2 is a technical revision of, and supersedes,
CP3:Chapter V:Part 2:1972. Both wind codes are currently in use, but the intention is to
withdraw CP3:Chapter V: Part:2 1972.
Dead loads are loads which are due to the effects of gravity, i.e. the selfweight of all
permanent construction such as beams, columns, floors, walls, roof and finishes etc.
If the position of permanent partition walls is known their weight can be assessed and
included in the dead load. In speculative developments, internal partitions are regarded
as imposed loading.
Imposed loads are loads which are due to variable effects such as the movement of
people, furniture, equipment and traffic. The values adopted are based on observation
and measurement and are inherently less accurate than the assessment of dead loads.
Clause 4. 0 and Tables 5 to 12 define the magnitude of uniformly distributed and
concentrated point loads which are recommended for the design of floors, ceilings and
their supporting elements. Loadings are considered in the following categories:
Residential:
Type 1 Selfcontained dwelling units
Type 2 Apartment houses, boarding houses, guest houses, hostels, lodging houses,
residential clubs and communal areas in blocks of flats.
Structural Steelwork 7
Most floor systems are capable of lateral distribution of loading and the recommended
concentrated load need not be considered. In situations where lateral distribution is not
possible the effects of the concentrated loads should be considered, with the load applied at
locations which will induce the most adverse effect, e.g. maximum bending moment, shear
and deflection. In addition, local effects such as crushing and punching should be considered
where appropriate.
In multistorey structures it is very unlikely that all floors will be required to carry the
full imposed load at the same time. Statistically, it is acceptable to reduce the total floor
loads carried by a supporting member by varying amounts depending on the number of
floors or floor area carried. This is reflected in Clause 5.0 and Tables 2 and 3 of
BS6399:Part 1, in wh a percentage reduction in the total distributed imposed floor loads
is recommended when designing columns, piers, walls, beams and foundations. It should be
noted that the loadings given in Table 3 do not include the imposed load carried by the roof.
Imposed loading caused by snow is included in the values given in this part of the code
which relates to imposed roof loads. Flat roofs, sloping roofs and curved roofs are also
considered. This part of BS 6399 is a draft version for public comment and should not be
used as a British Standard.
Environmental loading such as wind loading is clearly variable and its source is beyond
human control. In most structures the dynamic effects of wind loading are small and static
methods of analysis are adopted. The nature of such loading dictates that a statistical
approach is the most appropriate in order to quantify the magnitudes and directions of the
related design loads. The main features which influence the wind loading imposed on a
structure are:
Tabulated procedures enable these features to be evaluated and hence produce a system
of equivalent static forces which can be used in the analysis and design of the structure.
The application of the load types discussed in Section 1.2.2 to structural frames results in
axial loads, shear forces, bending moments and deformations being induced in the floor/roof
slabs, beams, colunms and other structural elements which comprise a structure. The
primary objective of structural analysis is to determine the distribution of internal moments
and forces throughout a structure such that they are in equilibrium with the applied design
loads. There are a number of mathematical models, some manual and others computer
based, which can be used to idealise structural behaviour. These methods include: two and
threedimensional elastic behaviour, elastic behaviour considering a redistribution of
moments, plastic and nonlinear behaviour. Detailed explanations of these techniques can be
found in the numerous structural analysis text books which are available; they are not
explained in this text.
In braced structures where floor slabs and beams are considered to be simply supported,
vertical loads give rise to three basic types of beam loading condition:
Consider the floor plan shown in Figure 1.3 where two oneway spanning slabs are
supported on three beams AB, CD and EF. Both slabs are assumed to be carrying a
uniformly distributed design load of 5 kN/m2 .
Structural Steelwork 9
E F E F
.
I 1.5 m
] 5 kN/m 2 3.0 m t
1.5 m
5 kN/m 2
~ 5.0 m
c Dj
2.5 m
t
AJ 8.0 m JB
J A
2.5 m
_j
8
(a) (b)
Figure 1.3
Both slabs have continuous contact with the top flanges of their supporting beams and
span in the directions indicated. The floor area supported by each beam is indicated in
Figure 1.3 (b).
BeamAB
Total load = (floor area supported x magnitude of distributed load/m2)
= (2.5 X 8.0) X (5.0) = 100 kN
Total Load
Beam CD r;;
Total load
BeamEF
= (4.0 X 8.0) X (5.0) = 160kN
Af ~B
c D
Total load (1.5 X 8.0) X (5.0) 60kN E F
Note: Beams AC, CE, BD and DF are assumed to be supporting only their selfweight in
addition to loading which may be applied directly such as brickwork walls.
Consider the same floor plan as in Example 1.1 but in this case the floor slabs are twoway
spanning as shown in Figure 1.4.
Since both slabs are twoway spanning, their loads are distributed to supporting beams
on all four sides assuming a 45° dispersion as indicated in Figure 1.4(b).
10 Design ofStructural Steelwork
+
E
5kN/m 2
+
c
5kN/m 2
BeamAB
js
68.75 kN
Beam EF
Tc
31.25 kN D 8
Beams CE and DF
=: 11.25 kN D F
The loading on beam CD can be considered to be the addition of two separate loads, i.e.
Load due to slab ABCD == 68.75 kN (as for beam AB)
Load due to slab CDEF == 48.75 kN (as for beam EF)
E 68.75 kN 5 48.75 kN
Ts + El tF
Structural Steelwork 11
Check: the total load on all beams 2(68 .75 + 48 .75 + 31.25 + 11.25)
320kN
Consider the same floor plan as in Example 1.1 with the addition of a secondary beam
spanning between beams AB and CD as shown in Figure 1.5. The load carried by this new
beam imposes a concentrated load at the midspan position ofboth beams AB and CD.
E F E F
~~~r~~r~~r~ r
5 kN/m 2
I 3.0m t
1.5 m
~
1.5 m
~~~~T~~~~~~ __l
c G C D
5 kN/m 2 5 kN/m 2
·  ·  
J
5.0m
o _m__
A ll_ 4_ ._ Ha_~_o_m_ 4_.o_m_il B
Figure 1.5
Beam EF
Total load (1.5 X 8.0) X (5.0) c 60kN
60kN
ET TF
BeamGH
c
Gr
100 kN
Total load (4.0 X 5.0) X (5 .0)
= 100 kN
lH
Beams AC and BD
Total load (2.0 X 5.0) X (5 .0)
c 50kN
50kN AT Tc
B D
Bean1AB l5o kN
Total load = End reaction from beam GH
= 50 kN AT Ts
Bean1 CD
The loading on beam CD can be considered to be the addition of two separate loads, i.e.
12 Design ofStructural Steelwork
c 60kN 150 kN
Considering the floor plan shown in Figure 1.6, with the oneway and twoway spanning
slabs indicated, determine the type and magnitude of the loading on each of the supporting
beams .
4.0 m 2.0 m F E
'I'
5 kN/m 2
IH
5 kN/m 2
J I
5 kN/m 2
'1'1 ~~~~r.r~~~ .
1.5 m
t
~I I~ 3.0 m
l
c G
~· c
+ 5 kN/m 2
5.0 m
J
AI 8.0 m 18 AI 2.5 m 3.0 m
I
Figure 1.6
The loads on beams AB, AC and.BD are the same as in Example 1.2.
= 15 kN
~c
D
H
J
BeamEF
End reaction from beam GH = 7.5 kN
End reaction from beam IJ 7.5 kN
Structural Steelwork 13
5 68 .75 kN
To
Total loads on beam CD due to beams GH, IJ and slab GHIJ and ABCD
7.5kN± 17.5kN
30 kN ... [68 .75 kN
cT To + c f To
The synthesis of a particular design solution evolves from the interaction of a wide range of
considerations to ensure that a structure does not become unsatisfactory in use during its
intended lifetime. The concepts of limit state design examine conditions which may be
considered to induce failure of a structure; such conditions are known as limit states. In BS
5950:Part 1: 1990 Structural use of steelwork in building two types of limit state are
classified, they are:
This is a consideration of the strength and/or stability of a structure and its component parts
at the onset of collapse and includes effects such as yielding, rupture, buckling, local and
overall instability, fatigue and brittle fracture.
This is a consideration of the response of a structure and its component parts while in
service and includes effects such as deflection, vibration, corrosion or any behaviour which
renders the structure unsuitable for its intended purpose when subjected to service loads.
The general principles of limit state design as applied to the structural steelwork design of
buildings are given in Section two of BS 5950:Part I.
The application of the partial safety factors to applied loading, in addition to BS6399:Parts
I and 2 referred to in section 1.2.2, is illustrated in Example 1.5.
Structural Steelwork 15
A partplan of a brewery building is shown in Figure 1.7. Using the data provided and the
relevant BS codes, determine the design shear force and bending moment for beams CD, EF
andAB.
A D F
225 mm thick
40 m3 Storage
brickwork wall
area for Brewers 5.0 m
3.5 m high
grain (wet)
I
~
2.2 X 225
Selfweight of 225 mm thick brickwork = 4.95 kN/m2
100
BS 5950:Part 1 Table 2
Partial safety factor for dead loads = 1.4
Partial safety factor for imposed loads = 1.6
In all three cases the critical load combination is the combined dead and imposed loading.
The design loads can therefore be determined by adding the factored dead and imposed load,
I.e.
Design load = (1.4 x dead load)+ (1.6 x imposed load)
BS 6399:Part 1: 1984
Table 10: Imposed load in motor room = 7.5 kN/m2
16 Design ofStructural Steelwork
100 kN
50kN~50kN
Design shear force = 50kN
WL (100 X 3.0) =
Design bending moment 37.5 kNm
8 8
BeamEF
Design dead load due to selfweight of slab = 1.4 X 3.6 X 1.5 = 7.56 kN
Design dead load due to brickwork wall = 1.4 X 4.95 X 10.5 = 72.77 kN
Design imposed load due to general office = 1.6 x 7.5 x 1.5 = 18.0 kN
Total design load = (7.56 + 72.77 + 18.0)
= 98.3 kN
98.3 kN
49.1kN~ 49.1kN
WL = (98.3 X 3.0)
Design bending moment =  36.9 kNm
8 8
BeamAB
Design dead load due to selfweight of slab in grain storage area
= 1.4 X 3.6 X 2.5 X 8.0 = 100.8 kN
Design dead load due to brickwork wall = 1.4 X 4.95 X 8.0 X 3.5 = 194.0 kN
Design imposed load due to wet grain = 1.6 X 110.0 = 176.0 kN
Total design load = (100.8 + 194.0 + 176.0)
= 470.8 kN
Structural Steelwork 17
320.8 kN
285.65 kN 344.5 kN
Shear Force Diagram
Note:
The value of the UDL between D and F= 95 .4 + 470 ·8 = 77.93 kN/m
5 8
153.1
x = = 1. 96 m this is the position of the maximum bending moment
77.93
Design bending moment is equal to the shaded area in the shear force diagram.
Two methods of analysis for determining the equivalent static wind loads on structures are
given in BS 6399:Part 2: 1995, they are:
A closed top storage hopper as shown in Figure 1.8 is situated in an industrial development
near Edinburgh and adjacent to the sea. Assuming the altitude of the location to be 5. 0 m
above mean sea level, determine the overall horizontal wind loading on the structure, while
considering the wind to be acting in the direction indicated.
8.0 m
Wind
direction
r1l Wi nd
direction 1i 0IT
8.0m
..r:::===h
1
Elevation
Figure 1.8
Solution:
Clause 2. 2. 3. 2
Since the crosswind breadth (8 .0 m) is less than the height {30.0 m) a reduction in lateral
loading is permitted.
Figure 11 (c) of BS 6399:Part 2 H =30m, B = 8.0 m :. H > 2B
Structural Steelwork 19
Consider the building surface to be divided into a number of parts A, B, C and D
r
B.Om
t Area= 64m2
I
Area= 56m2
I
Area= 56m 2
I
Area= 64m 2
Figure 1.9
Clause 2.1.3. 6
The overall load on the building P = 0.85(LPrront LPrear) (1 + Cr)
where:
LPrront is the sum of the horizontal components of surface load on the windward facing
wall
is the sum of the horizontal components of surface load on the leeward facing wall
is a dynamic augmentation factor
Clause 2.1.3.5
Net load on an area of surface P=pA
where:
p is the net pressure across the surface
A is the area of surface being considered
Clause 2. 2. 2. 2. 2 Figure 7
Assuming that the topography of the site is not considered significant
altitude factor S. = 1 + 0.001L\s
1 + (0.001 X 5)
s. = 1.005
Clause 2. 2. 2
site wind speed v.
= vb Xs. Xsd Xs. XSp
In many cases the direction factor (Sct), seasonal factor (S.) and probability factor (Sp) can
be considered to be equal to 1.0 (see Clauses 2.2.2.3 to 2.2.2.5).
V';; = (23.5 X 1.005)
= 23.62 m/sec
20 Design ofStructural Steelwork
Clause 2. 2. 3
effective wind speed Ve = v.
Xsb
where:
Sb is the terrain and building factor obtained from Clause 2. 2. 3. 3 and Table 4
Table 4
Area A He= 8.0m sb = 1.73
AreaB He= 15.0 m sb = 1.85
AreaC He=22.0m sb = 1.91
AreaD He= 30.0 m sb = 1.96
Effective wind speeds
Area A Ve = 23.62 x 1.73 = 40.9 m/sec
AreaB Ve = 23.62 x 1.85 = 43.7 m/sec
AreaC V., = 23.62 x 1.91 = 45.1 m/sec
AreaD Ve = 23.62 x 1.96 = 46.3 m/sec
Clause 2.1.3.1
External surface pressure Pe = q.CpeCa
where:
Cpe is the external pressure coefficients (Clause 2.4)
Ca size effect factor (Clause 2.1. 3.4)
Note: In this problem it is not necessary to consider internal pressure coefficients since the
overall horizontal loading is being considered. Internal pressure coefficients are considered
in Example 1. 7.
In this problem the wind loading on the side faces and roof are not being considered.
Clause 2.1.3.5
Net load on building surface area = p x loaded area
Area A Frront = 0.73 X 64= + 47.72 kN Frear 0.28 X 64 = 17.92 kN
AreaB Prront 0.83 X 56=+ 46.48 kN Frear 0.31 X 56 17.36kN
AreaC Prront = 0.89 X 56=+ 49.84 kN Frear 0.33 X 56  18.48 kN
Area A Prront = 0.93 X 64= + 59.52 kN Frear = 0.35 X 64 =  22.40 kN
Clause 2.1. 3. 6
The overall horizontal load on the buildingF = 0.85(~rront ~rear) (1 + Cr)
Design Data:
Location open country near Preston
Altitude of site 20.0 m above mean sea level
Closest distance to sea 8.0km
Overall length of building 24.0m
Centres of frames 4.0m
Frames at 4.0 m
centres Pinned base
"20.0 m   1
Figure 1.10
Solution:
Clause 2.2.3.2
Figure 11 (a) H = 10 m B=24m H < B
.·. consider building as one part
Clause 2.1.3.2
Internal surface pressures Pi
D = 20 = 2 _0 20.0 m
H 10
U wind direction
Figure 1.11
Clause 2. 4 Table 5
The values of Cpe in Table 5 can be interpolated in the range 1 < D < 4
H
i 0.2
leeward face
windward face
Cpe = 0.23
Cpe = + 0.73 D i 0.7
Figure 1.12
In the case of the gable faces, different values of Cpe should be used depending on the
gap between adjacent buildings; in this example assume that the building is isolated.
24 Design ofStructural Steelwork
Clause 2.4.1.3 Figure 12
The surfaces on which the wind is blowing are considered as separate zones which are
defined in Figure 12 and depend on a variable 'b ',
where:
b ::=:; crosswind breadth = 24.0 m
::=:; 2 x height = 2 x 10.0 = 20.0 m
Use the smaller value of b i.e. b = 20.0m
zone zone
A B
4m 6m Figure 1.13
q 20.0 m
wind direction
D 24
==2.4
H 10
6 bays at 4.0 m =24.0 m
Figure 1.15
Structural Steelwork 25
windward face
leeward face
Cpe
Cpe
= +0.71
= 0.21 0.71 __. 0 _. 0.21
Figure 1.16
Clause 2. 4.1. 3
b s crosswind breadth = 20.0 m
s 2 x height 2 x 10.0 = 20.0 m
Use the smaller value of b 1.e. b 20.0 m
Figure 12(b) D = 24.0 > b :. consider the longitudinal surface as three zones A, Band C.
Longitudinal elevation
Figure 1.17
Clause 2.5.2.2 this clause defines the loaded zones which are indicated in Figure 20 and
are based on variables bL and bw
where:
bL s L (crosswind dimension with the wind on the side of the building)
s 2H
26 Design ofStructural Steelwork
G 8.0 m
E I Fl E 2.0m
c 8.0 m
A IB I A 2.0 m
I I
10.0 m 4.0 m 10.0 m
wind direction e = 0°
Plan
Figure 1.19
Both +ve and ve values are given for zones A, B and C, the most onerous value should
be selected when considering combinations with internal pressure coefficients and other load
types.
Figure 1.20
Structural Steelwork 27
The zones A, B, E and Fare normally used when designing for local effects where high
local suction can occur. When calculating the load on entire structural elements such as
roofs and walls as a whole, then the values for C and G should be adopted as shown in
Figure 1.21
Figure 1.21
A
5.0 m
r
f c D
wind 5.0 m 8
~
~~ c
direction 5.0 m
D
8 = 90° 5.0 m A
~
Figure 1.23
As before use zones A and B for local effects and zone C which is more onerous than D
for entire structural elements.
porosity to the internal volume containing the opening. In cases when dominant openings
occur they will control the internal pressure coefficients and should be determined using
Table 17; in other cases Table 16, Clauses 2. 6.1.1 and 2. 6.1. 2 should be used.
In many cases for external walls Cpi should be taken as either 0.3 or +0.2, whichever
gives the larger net pressure coefficient across the walls.
I.e.
0 .3 0 .2
+ +
Figure 1.24
rn rn
1.3 0.8 1.3 0.8
t t t t
0.73 0.23 0.73 0.23
c::::> + + q + +
wind direction wind direction
wind angle e = 0° ~ ~ wind angle e = 0° ~ ~
Figure 1.25(a)
0.45
0 .73 0 .45
+ + 0 .2 + +
Case II
Figure 1.25(b)
Structural Steelwork 29
rn+
0.2 0.2
t t
0.21+ 0.4 0.21+ rn+ 0.4
0.8 + + 0.3 + + 0.8 +
0.8 + + 0.2 + 0.8
1.3 + t + 1.3 1.3 + ~ + 1.3
t t
0.71 0.71
wind direction l'f' wind direction l'f'
wind angle 8 = 90° U wind angle 8 = 90° U
~I +
~
0.3 ... lo~ ~I + 0.2
+ + lo~
Caselli Case IV
Figure 1.26
Consider Case I
Clause 2.1.3.1 and 2.1.3.3
I I
r
0.95 kN/m2 0.03 kN/m2
wind==::::::::>
1
direction '·:   :      l
wind angle e = 0° J,. J,.
1.02 kN/m2 0.5 3kN/m2
zone A zone B
Figure 1.27
Clause 2.1.3.5
The net load on an area of building is given by P = pA
A typical internal frame supports surface areas as shown in Figure 1.29
Figure 1.28
Only Case I is considered here; when designing such a frame all cases must be
considered in combination with dead and imposed loads and appropriate partial load factors
as given in Table 2 ofBS 5950:Part 2, to determine the critical design load case.
32 Design ofStructural Steelwork
22.8 kN
Figure 1.29
Data:
Location near Aberdeen
Closest distance to sea 5km
Altitude above mean sea level 15m
r1 0.0 m +
··r·.· ··········.~:?"<:~..·········· Jf m
Figure 1.30
Solution:
Clause 2.1.3.3 (b)
When considering free standing canopies and building elements, the net surface pressure is
given by
where CP is the net pressure coefficient for a canopy surface or element and is defined in
Clauses 2.5.9 and 2. 7. In this problem the radar equipment can be treated as indicated in
Clause 2. 7. 6; the wind force on the members of the trestle will be neglected.
:. Cp = 1.8
Clause 2.1.3.4 Figure 4
Size effect factor (C.) He= 8.0 m, Site in country, Distance to sea = 5 km
Structural Steelwork 33
Use line A in graph assuming the diagonal dimension 'a' to be 5. 0 m
c.~ 0.99
Clause 2. 2. 2. 2. 2
Altitude factor Sa = 1 + (0.001 X 15)
= 1.015
Clause 2. 2. 2
Site wind speed v. = vb X s. X sd X s. X Sp
Assume Sd = s. = SP = 1.0
V. = 24.0 x 1.015 = 24.36 m/sec
Clause 2.2.3.3 Table 4
The values for Sb in Table 4 can be found by interpolation.
In this example He= 8.0 m and closest distance to sea is 5 km :. Sb = 1.68
Clause 2.2.3.1
Effective wind speed Ve = 24.36 X 1.68 = 40.9 m/sec
Clause 2.1. 2.1
Dynamic wind pressure q. (0.613 x 40.9 2)/10 3 = 1.03 kN/m2
Net surface pressure p = 1.03 X 1.8 X 0.99 = 1.84 kN/m2
Clause 2.1.3.5
Net load on the reflector p = 1.84 X (10.0 X 4.0) = 73.6 kN
This force should be considered to be acting at the midheight of the reflector and within the
middle half as shown in Figure 1.31.
Figure 1.31
2. Flexural Members
2.1 Introduction
The most frequently used, and possibly the earliest used, structural element is the beam. The
primary function of a beam is to transfer vertical loading to adjacent structural elements
such that the load can continue its path through the structure to the foundations. Loading
can be imposed on a beam from one or several of a number of sources, e.g. other secondary
beams, columns, walls, floor systems or directly from installed plant and/or equipment. In
most cases static loading will be considered the most appropriate for design purposes, but
dynamic and fatigue loading may be more critical in certain circumstances.
The structural action of a beam is predominantly bending, with other effects such as
shear, bearing and buckling also being present. In addition to ensuring that beams have
sufficient strength capacities to resist these effects, it is important that the stiffuess
properties are adequate to avoid excessive deflection or local buckling of the crosssection
(see Section Classification 2.2). A large variety of crosssections are available when
selecting a beam for use in any one of a wide range of applications. The most common types
of beam with an indication of the span range for which they may be appropriate are given in
Table 2.1. For lightly loaded and small spans such as roofpurlins and side sheeting rails the
use of hotrolled angle sections or channel sections is appropriate (see Figure 4.11). Cold
formed sections pressed from thin sheet and galvanised and provided by proprietary
suppliers are frequently used. In small to medium spans hotrolled joists, universal beams
(UBs ), hollow sections and UBs with additional welded flange plates (compound beams) are
often used. If the span and/or magnitude of loading dictates that larger and deeper sections
are required, castellated beams formed by welding together profiled cut UB sections, plate
girders or box girders in which the webs and flanges are individual plates welded together
can be fabricated. Plate girder design is discussed in more detail in Chapter 6. While careful
detailing can minimize torsional effects, when they are considered significant hollow tube
sections are more efficient than open sections such as UBs, universal columns (UCs), angles
and channels.
The section properties of all hotrolled sections and coldformed sections are published
by their manufacturers; those for fabricated sections must be calculated by the designer
(Ref.: 11)
The span of a beam is defined in Clause 4.2.1.1 of BS 5950:Part 1 as the distance
between points of effective support. In general, unless the supports are wide columns or
piers then the span can be considered as the centretocentre of the actual supports or
columns.
The most widely adopted section to be found in building frames is the Universal Beam.
The design of beams to satisfy the requirements of BS 5950:Part 1 includes the
consideration of:
+ section classification,
+ shear capacity,
34
Flexural Members 35
All of these criteria are explained and considered separately and illustrated in Examples 2.1
to 2. 10. Examples 2.11 and 2.12 illustrate the design of beams using all criteria (except
torsional capacity).
Table 2.1
1  20
L[ID
Angle Channel Joist Tube Universal beam Compound beam
140
15  200
K}{8}0( I \: :::::», 7
Castellated beams Welded plate girders Welded box girders
(i) the aspect ratio of the elements of a crosssection such as the UB indicated in Figure
2.1, which influences their behaviour when subject to either pure compression,
compression caused by bending or a combination of both.
b
I· 1_1
T Element Aspect ratio
D j~ outstand of
compressiOn
flange
biT
web d/t
I B I
Figure 2.1
Class 1
Class 2
Class 3
Rotation
Figure 2.2
where:
M=p plastic moment of resistance
Me= limiting elastic moment of resistance
M = elastic moment of resistance
These criteria detennine whether or not a fully plastic moment as shown in Figure 2.3 can
develop within a beam and possess sufficient rotational capacity to permit redistribution of
the moments in a structure.
In a beam subject to an increasing bending moment, the bending stress diagram changes
from a iinearly elastic condition with extreme fibre stresses less than the design strength,
Flexural Members 37
(py), to one in which all of the fibres can be considered to have reached the design strength
as shown in Figure 2.3.
Compression cr <py (j = py cr = py
~~L~
Tension (j <py cr = py (j = py
Elastic Limiting Fully Plastic
Moment Elastic Moment Moment
Figure 2.3
where:
Zxx elastic section modulus
Sxx = plastic section modulus
cr elastic stress
py design strength
plastic modulus
v= Sxx The value of v for most Isections ~ l. 15
elastic modulus Zxx
The failure of a structure such that plastic collapse occurs is dependent on a sufficient
number of plastic hinges developing within the crosssections of the members (i.e. value of
internal bending moment reaching Mp), to produce a mechanism. For full collapse this
requires one more than the number of redundancies in the structure, as illustrated in Figure
2.4.
38 Design ofStructural Steelwork
p w p
/ Ama:l ~
Figure 2.4
The required number of hinges will only develop if there is sufficient rotational capacity
in the crosssection to permit the necessary redistribution of the moments within the
structure. When this occurs, the stress diagram at the location of the hinge is as shown in
Figure 2.3(c), and the aspect ratio of the elements of the crosssection are low enough to
prevent local buckling from occurring. Such crosssections are defined as plastic sections
and classified as Class 1. Full plastic analysis and design can only be carried out using
Class 1 sections.
When crosssections can still develop the full plastic moment as in Figure 2.3(c), but are
prevented by the possibility of local buckling from undergoing enough rotation to permit re
distribution of the moments, the section is considered to be compact and is classified as
Class 2. Compact sections can be used without restricting their capacity, except at plastic
hinge positions.
Semicompact sections may be prevented from reaching their full plastic moment by local
buckling of one or more of the elements of the crosssection. The aspect ratios may be such
that only the extreme fibre stress can attain the design strength before local buckling occurs.
Such sections are classified as Class 3 and their capacity is therefore based on the
limiting elastic moment as indicated in Figure 2.3(b).
When the aspect ratio is relatively high, then local buckling may prevent any part of the
crosssection from reaching the design strength. Such sections are called slender sections
and are class~fied as Class 4 sections; their capacity is based on a reduced design
strength as spec~fied in Clause 3. 6 of BS 5950: Part 1.
Flexural Members 39
The limiting aspect ratios for elements of the most commonly used crosssections subject
to pure bending, pure axial load or combined bending and axial loads are given in Table 7
( r.
of BS 5950:Part 1. The values given in Table 7 must be modified to allow for the design
strength py. This is done by multiplying each limiting ratio by E which equals
~: The usc of Clause 3. 5 and Table 7 is illustrated in the examples in this chapter.
Generally in the design of beams for buildings, the effects of shear are negligible and will
not significantly reduce the value of the moment capacity. It is evident from the elastic shear
stress distribution in an 1beam, as shown in Figure 2.5 that the web of a crosssection is the
primary element which carries the shear force.
!
Applied shear force Maximum shear stress
I\._
Applied loading ,
, 6;{6;r     
I
/r
\~Q
I
c,,, '
\
:' ',/ /Q',
1 7 r
I
'
\
' : l · I ,___1 jof
I Ir / I '
I \
'1 I
\
',lT_\ ~ }'u/
r /
Figure 2.6
40 Design ofStructural Steelwork
where:
0. 6 py is approximately equal to the yield stress of steel in shear
Av is the shear area as defined in Clause 4. 2. 3
When the applied ultimate shear force (Fv) is equal to or greater than 60% of Pv (i.e. Fv
~ 0.36 pyAv) the moment capacity of a beam should be reduced as specified in Clause 4.2.6.
When the aspect ratio of a web is greater than 63s the possibility of shear buckling
should be considered. This is illustrated in the design of plate girders in Chapter 6. In the
design of webs of variable thickness and/or which contain large holes (e.g. castellated
beams) the code requires that shear stresses be calculated from first principles assuming
elastic behaviour and a maximum shear stress not exceeding 0.7py; the design of such webs
is not considered in this text.
A simply supported 406 x 178 x 74 UB is required to span 4.5 m and carry an ultimate
design load of 40 kN/m. Check the suitability of the section with respect to shear (section
properties are given in Ref: 11).
Solution:
Section properties: t = 9.5 mm, D = 412.8 mm, d = 360.4 mm
Design shear force at the end of the beam Fv = 40 x 4 .5 = 90 kN
2
Clause 4.2.3 Pv = 0.6 pyAv
For a rolled UB section Av = tD
Shear area Av = (9.5 x 412.8) = 3.922 x 103 mm2
Clause 3.1.1 Web thickness t = 9.5 mm
Table 6 gives py= 275 N/mm2
Clause 3.5 Since the beam is subject to pure bending the neutral axis will be at mid
depth.
Section Classification:
d
Table 7 Note: s = 1.0 = 37.9 < 79s
t
Web is plastic
.
Sl1ear capactty p 0.6 X 275 X 3.922 X 103
v= 3 = 64 7 kN
10
>> Fv (90 kN)
This value indicates the excessive reserve of shear strength in the web.
Flexural Members 41
2.4 Moment Capacity (Clause 4.2.5)
The criteria (i) to (v) are relatively straightforward to evaluate, however criterion (vi) is
related to the lateral torsional buckling of beams and is much more complex. The design of
beams in this text are considered in two categories:
(a) beams in which the compression flange is fully restrained and lateral torsional
buckling cmmot occur, and
(b) beams in which either no lateral restraint or only intermittent lateral restraint is
provided to the compression flange.
As indicated in Figure 2.3 a beam subject to bending is partly in tension and partly in
compression. The tendency of an unrestrained compression flange in these circumstances is
to deform sideways and twist about the longitudinal axis as shown in Figure 2.7.
Figure 2.7
This type of failure is called lateral torsional buckling and will normally occur at a value
of applied moment less than the moment capacity (Me) of the section as given in Clauses
4. 2. 5and 4. 2. 6and is known as the buckling resistance moment Mb, defined in Clause
4.3. 7. 3 . as :
where:
Sxx. is the plastic modulus
Pb is the bending strength
42 Design ofStructural Steelwork
The tendency for the compression flange to deform is influenced by:
Full lateral restraint is defined in Clause 4.2. 2. as being present '... if the frictional or
positive connection of a floor or other construction to the compression flange of the
member is capable of resisting a lateral force of not less than 2.5% of the maximum
factored force in the compression flange of the member under factored loading. This load
should be considered as distributed uniformly along the flange ... '. This is illustrated in
Figure 2.8.
Figure 2.8
Flexural Members 43
where:
M maximum moment due to applied factored loads
L span of the beam
f.1 coefficient of friction between the concrete beam and the steel flange
R maximum factored verticalload/m applied to the beam
D and T are as before
As discussed above, there is no need to carry out this calculation in the case of concrete
floor slabs.
Effective length of
compression nange
Secondary beam
providing lateral and
torsional restraint
Secondary beam
providing lateral and
torsional restraint
Figure 2.9
It is important to ensure that the elements providing restraint are an integral part of a
braced structural system and are capable of transmitting the lateral force of 2.5% described
previously, divided between the intermediate lateral restraints in proportion to their spacing
(see Clause 4. 3.2.1). If three or more intermediate lateral restraints are present, each
individual restraint must be capable of resisting at least I% of the flange force.
44 Design ofStructural Steelwork
2.4.1.2 Torsional restraint (Clause 4.3.3)
A beam is assumed to have torsional restraint about its longitudinal axis at any location
where both flanges are held in their relative positions by external members during bending;
as illustrated in Figure 2.10.
Figure 2.10
This type of restraint may be provided by load bearing stiffeners as described in Clause
4. 5. 8 or by the provision of adequate end connection details as discussed in Chapter 5.
Figure 2.11
The provision of lateral and torsional restraints to a beam introduces the concept of
effective length. The effective length of a compression flange is the equivalent length
between restraints over which a pinended beam would fail by lateral torsional buckling.
The values to be used in assessing this are given in Tables 9 and 10 for beams and
cantilevers respectively. The values adopted depend on three factors relating to the degree of
lateral and torsional restraint at the position of the intermittent restraints, they are:
l'lexura/A1e1nbers 45
In the case of beams factors (a) and (b) give rise to five possible conditions.
(b) When both flanges are free to rotate on plan and the compression flange is
unrestrained:
(i) torsional restraint is provided soley by connection of the tension flange
to the supports,
(ii) torsional restraint is provided soley by dead bearing of the tension flange
on supports.
Crab
~V•rtl~l Lood ~
Crane Gantry
Gl'd"
=I'
CJ
Figure 2.12
46 Design ofStructural Steelwork
In Table 10 it can be seen that destabilising loads are a particular problem for
cantilevers since it may be difficult to achieve torsional rigidity at either the free or fixed
end. In addition the bottom flange, which is in compression, may not be as readily restrained
as the top flange.
The moment capacity (Me) of beams with the compression flange fully restrained is
detem1ined using the following equations which are given in Clauses 4. 2. 5 and 4. 2. 6 for low
coincident shear and high coincident shear respectively.
The design of hollow sections, (which tend to have a high shear capacity and high
torsional stiffness and do not generally fail by lateral torsional buckling), will be inefficient
and uneconomic if this enhancement is not allowed for.
Sv is either
(i) the plastic modulus of the shear area for sections with equal flanges or
(ii) plastic modulus of the gross crosssections minus the plastic modulus of that part
of the section remaining after deduction of the shear area.
The shear areas in (i) and (ii) are illustrated in Figure 2.12.
i'lexuralA1e~nbers 47
0 d d
Figure 2.12
Se1nico~npact sections
Low Shear and High Shear
Clauses 4. 2. 5 and 4. 2. 6
The same value is used for both situations i.e.:
Slender sections
Low Shear and High Shear
Clauses 4. 2. 5 and 4. 2. 6
The same value is used for both situations i.e.:
where Pr is multiplied by a factor from Table 8 to allow for the slenderness of the cross
section.
The design of beams with fully restrained flanges is normally determined from the
bending criterion with subsequent checks carried out to ensure that other criteria are
satisfied. A beam can be selected such that the applied ultimate moment does not exceed the
moment capacity of the section, e.g. for a plastic or compact section assuming that the
shape factor is less than 1.2:
A single span beam is simply supported between two columns and carries a reinforced
concrete slab in addition to the column and loading shown in Figure 2.13. Using the
working loads indicated, select a suitable section considering section classification, shear
and bending only. Assume that dead loads are inclusive of selfweights.
48 Design of Structural Steelwork
Dead load = 10 kN
Imposed load= 30 kN !I

Dead load = 8 kN/m
lmpoj ed load = 10 kN/m
i I~
I[ I
i IT
.L: I
2.0m
I 4.0m
i
 I
6.0m
.I
,..,....
Figure 2.13
2.6 Moment Capacity (Mb) of Beams without Full Lateral Restraint (Clause 4.3. 7)
The bending strength Pb is dependent on the design strength py the equivalent slenderness
ArT (equal to nuvA.which is defined in Example 2.3) ofthe section and the type of member,
i.e. rolled or fabricated by welding. Values of Pb are given in Tables 11, 12 and 19 of the
code. Those in Tables 11 and 12 are for rolled and welded sections respectively and those in
Table 19 are used when using a simplified design procedure which is known as the
Conservative Method. A more rigorous and economic design process is illustrated first in
this Section. The conservative method and the use of safe load tables are illustrated at the
end of the chapter.
where:
M is the equivalent uniform moment on any portion of a member between adjacent
lateral restraints and is defined in Clause 4. 3.7. 2 as mMA in which:
JllexuralAfe111bers 49
A single span beam of 8. 0 m span supporting two factored point loads is shown in Figure
2.14. Assuming lateral and torsional restraint to the compression flange at the ends and
points of application ofthe loads only, check the suitability of a 406 x 140 x 39 UB with
respect to bending. Neglect selfweight, i.e. the beam is not loaded between adjacent lateral
restraints.
45 kN
30 kN
Figure 2.14
The evaluation of 'AlA' is dependent on the structural analysis for the applied load
system and resulting bending moment diagram.
Clause 4.3. 7.6: This clause indicates two situations for which to determine '111'
Table 13: 'm' is given the value of 1.0 for any member subject to destabalising loads,
any member which is loaded between adjacent restraints and any section
with unequal flanges.
In the case of members with equal flanges not loaded between adjacent
lateral restraints and not subjected to destabalising loads the value of 'm' is
determined from Table 18.
The values of 'm' and 'n' as determined by Table 13 requirements are
to allow for the effects of a moment gradient along the length of the
member. It should be noted from Table 13 that the two adjustments, 'm' to
the applied bending moment and 'n' to the slenderness of the compression
flange do not occur simultaneously.
Table 18: the magnitude and sense of the bending moment diagram on the portion of
beam being considered defines the value of a factor 'p which is
subsequently used to determine 'm' where:
67.5 kNm
82.5 kNm
Figure 2.15
0 0
fJ=  = 0 fJ= 675 = 0.82 {J==0
67.5 82.5 82.5
Table 18 gives values of f3in increments ofO.l, the value of 'm' for intermediate values
can be found from the following equation but should not be taken as less than 0.43
Clearly, in the above case portion BC of the beam is critical and a section should be
designed such that:
Mb ;::: 75.1 kNm
ALT = nuvA,
where:
n is a slenderness correction factor and is equal to either 1.0, or for members with equal
flanges, loaded between adjacent restraints and not subjected to destabilising loads, can
be determined from Tables 15 and Table 16 (see Table 13 as form) or for standard load
conditions, from Table 20.
u is a buckling parameter which for rolled UB, UC or channel section can be taken as 0.9,
is more accurately evaluated using Appendix B, or is given in published section property
tables.
v is a slenderness factor which for uniform, flanged members with at least one axis of
symmetry can be determined from Table 14.
Evaluation of 'n'
In most cases, the value of 'n' can be estimated by modifying, when required, the values
given in Table 20. Tables 15 and 16 are used for more complex cases, as illustrated in
Example 2.4. The value depends on the bending moment diagram of the portion of a beam
being considered. Consider the portions of beam ABCD as before and using Table 20.
PortionAB
n = 0.77
67.5 kNm
Portion BC
(1.0 + 0.77)
Use average value :. n = = 0.89
2
52 Design of Structural Steelwork
Portion CD
82.5kNm V n = 0.77
N this factor equals 0.5 for members with equal flanges or in the case of other sections
/cf
equals =
fer+ ltJ:
where fer and ltr are the second moments of area of the compression and tension flanges
respectively about the yy axis of the section.
X is the torsional index and may be determined from Appendix B, section property tables
or taken as DIT for rolled sections, provided that U is taken as 0. 9 or 1. 0 for other
sections.
A 406 x 178 x 54 UB is simply supported and carries factored loading as shown in Figure
2.6. Assuming lateral restraints to the compression flange at A, B, C and D as in Example
2.3, check the suitability of the section with respect to bending.
45 kN
81 .75
27.75
X 20.25
.I
65.25
89.25
A B c D E
Figure 2.17
27.75
Distance to point of zero shear from B = 2.31 m
x =
12
Position of maximum bending moment is 4.31 m from the lefthand end. ·
54 Design ofStructural Steelwork
(81.75 + 57.75)2
Bending moment at B = 139.5 kNm
2
( 65.25 + 89 .25)2
Bending moment at C 154.5 kNm
2
(i) y=MIMo
where:
M is the greater end moment on the portion of beam being considered,
Mo is the midspan moment on a simply supported span equal to the unrestrained length.
The signs and values to be adopted forM and Mo are illustrated in Table 17.
A
Table 14 N=0.5 and  = 2.72 v = 0.92
X
Clearly, this section has insufficient bending capacity to support the load and either a
larger section or a higher grade of steel should be used.
Assume grade 50 steel
A single span beam supports the 'service' loads indicated in Figure 2.18 and is restrained at
points ABCD as in Examples 2.3 and 2.4. Check the suitability of a 300 x 200 x 8 RHS
section with respect to bending.
Figure 2.19
This clause states 'Box sections of uniform wall thickness need not be checked for lateral
torsional buckling effects provided that A, = (L/ry) is not greater than the limiting values
given in table 38 '; i.e. when
JflexuralA1e1nbers 57
DIB A,
1 00
350 X 275
2 Py
225 X 275
3 Py
170 X 275
4 Py
The limiting value of A, for DIB = 2.0 is given by 350 x 275 = 350
Py
The critical span above which this section must be checked is given by:
Lateral torsional buckling is clearly not critical and the moment capacity is governed by
Clause 4.2.5 (Note: low shear occurs in this case).
Afc 1.53pyZxx
(1.53 X 275 X 653 X 103)/106 ::::274.7 kNm
Consider a beam with a cantilever overhang and supporting the factored loads shown in
Figure 2.20.
58 Design ofStructural Steelwork
A maximum bending moment and a maximum shear force always occur at the same
crosssection in a cantilever. It is essential in these circumstances to ensure that a check is
carried out on the effect of the shear force on the moment capacity i.e. either Low Shear
with Fv < 0.6Pv in which case Clause 4.2.5 governs, or High Shear with Fv > 0.6Pv, in
which case Clause 4. 2. 6 is used.
Check the suitability of a 356 x 171 x 51 UB with respect to combined shear and
bending.
250 kN
104.2 kN
A c
43.24 kNm
Figure 2.21
At position B
Design shear force = 270.8 kN
Design bending moment = 240kNm
Section properties:
D = 355 mm B 171.5 mm T = 11.5 mm
ryy = 39.11 mm Sxx = 1010 X 103 mm3 = 7.4 mm
X = 28.8 Zxx 896 x 103 mm3 u = 0.882
Flexural Members 59
This UB section does not satisfy the lateral torsional buckling criterion. A larger section,
higher grade of steel, additional restraint to the compression flange or a combination of
these modifications can be used to solve this problem.
The very high value of loading in this problem demonstrates the reason for most beams
being designed for bending with coincident low shear. The design cases in which checks for
high shear are necessary are:
In addition to shear failure of a web as discussed in section 2.3, there are two other modes
of failure which may occur, they are:
At locations of heavy concentrated loads such as support reactions or where columns are
supported on a beam flange, additional stress concentrations occur in the web. This
introduces the possibility of the web failing in a buckling mode similar to a vertical strut, or
by localised bearing failure at the top of the root fillet, as shown in Figure 2.22.
Critical section
Critical section for
for web buckling
web bearing L
~Applied concentrated load  
Figure 2.22
Flexural Members 61
The code specifies two local capacities relating to these modes of failure. When either of
these is less than the applied concentrated force it will be necessary to provide additional
strength to the web. In most cases this requires the design of load bearing stiffeners; the
detailed design of such stiffeners is given in Chapter 6. There may be other reasons for
utilizing stiffeners, such as enhancing torsional stiffness at supports and points of lateral
restraint, as discussed previously.
In the buckling check the web is considered to be a fixedend strut between the flanges. It is
assumed that the flange through which the concentrated load is applied is restrained against
rotation relative to the web and lateral movement relative to the other flange.
The load carrying capacity of a strut is dependent on its compressive strength 'pc' and
crosssectional area and is given by:
Figure 2.23
LE = 0.7d r
YY
= ~ Area
1 (b1 +n1 )t 3
12(b1 +n1 )t = fu= t
2.f3
A = 0.7d r:::;
2.5d
(2~)
t
When a flange is not restrained against rotation and lateral movement, the value of A
should be modified by using an effective length based on the values given in Table 24.
The bearing check is similar to the buckling check, in that an effective bearing area over
which the design strength of the web is assumed to act is determined using:
where:
b, is as before
nz is the length obtained by dispersion through the flange to the top of the root fillet
assuming a slope 1:2.5 to the plane of the flange and
pyw is the design strength ofthe web
Figure 2.24
Flexural Members 63
In both cases at the design stage it is usually necessary to make assumptions regarding
the provision of bearing plates at supports or cap plates/base plates on columns to provide
stiff bearing. In the code, Clause 4.5 .1.3 defines the stiff bearing length b 1 as 'that length
which cannot deform appreciably in bending'. The value ofb 1 is determined by assuming a
dispersion ofload through a bearing plate at 45° and is illustrated in Figure 8 of the code.
Consider the beam in Example 2.2 in which the lefthand end reaction is 122.9 kN, and
check the suitability of the web with respect to buckling and bearing.
approx 0. 59r
90 X 90 X 10
Angle section u
T'
Figure 2.25 Web buckling
slope = 1:2.5
Web buckling:
Pw == (bi + n1)tpc
Assuming the bottom flange is laterally and torsionally restrained,
A 2.5d/t == (2.5x 265 .2) == 99
6.7
Table 6 < 16 mm py == 275 N/mm2 A == 99
Pc == 127 N/mm2
Web bearing:
Pcrip== (b1 + n2)tpyw
Pcrip== [(21.5 +52) X 6.7 X 275]/103 == 135 kN
Pcrip> 122.9 kN
Web bearing strength is adequate
A similar calculation can be carried out at the location of the column on the top flange.
62 kN
Figure 2.27
Since the load is normally less and the distribution in the web is considerably greater,
this will generally be less critical than the location at the end reaction. In the case of square
and rectangular hollow sections when the flange is not welded to a bearing plate, additional
effects of moments induced in the web due to eccentricity of loading as shown in Figure
2.28 must be allowed for.
Flexural Members 65
Bending is induced
in the webs
Figure 2.28
Reference should be made to the Structural Design Guide ... (Ref: 11, 12) which contains
detailed information relating to the relevant bearing factors in such cases.
In Table 1 of the code, one of the seniiceability limit states to be considered is deflection.
Recommendations for limiting values of deflection under various circumstances are given in
Table 5.
Limitations on the deflections of beams are necessary to avoid consequences such as:
There are large variations in what are considered by practising engineers to be acceptible
deflections for different circumstances. If situations arise in which a designer considers the
recommendations given in Table 5 to be too lenient or too severe (e.g. conflicting with the
specification of suppliers or manufacturers) then individual engineering judgement must be
used.
The values in Table 5 relating to beams give a span 1 ratio calculated using the
coefficient
service loads only. The coefficient varies from 180 for cantilevers, to 360 for the deflection
of beams supporting brittle finishes. In most circumstances, the dead load deflection will
have occurred prior to finishes being fixed and the building being in use and will not
therefore cause any additional problem while the building is in service. Unfactored loads are
used since it is under service conditions that deflection may be a problem.
1A ratio is used instead of a fixed value since this limits the curvature of the beam which depends
on the span.
66 Design ofStructural Steelwork
Additional values are given for portal frames and cranegantry girders; they are not
considered here.
In <1; simply supported beam, the maximum deflection induced by the applied loading
always approximates to the midspan value if it is not equal to it. A number of standard,
frequently used load cases for which the elastic deformation is required are given in Table
2.1 of this text. In the case indicated with '*' the actual maximum deflection will be
approximately equal to the value given (i.e. within 2.5%).
In many cases beams support complex load arrangements which do not lend themselves
to either an individual load case or to a combination of the load cases given in Table 2.1.
Since the values in Table 5 are recommendations for maximum values, approximations in
calculating deflection are normally acceptable. Provided that deflection is not the governing
design criterion, a calculation which gives an approximate answer is usually adequate. The
Steel Designers' Manual (Ref: 17) provides a range of coefficients which can be used either
to calculate deflections or to determine the minimum I value (second moment of area), to
. fy . span
sat1s any part1cu1ar . ratio.
coefficient
An equivalent uniformly distributed load technique which can be used for estimating
actual deflections or required I values for simply supported spans is given in this text.
Nonuniform loading
The equivalent UDL (We) which would induce the same magnitude of maximum bending
moment (note that the position may be different) on a simply supported span carrying a
un~form loading can be determined from:
8 B.Max.
We=
Table 2.1
Maximum Maximum
Load Case Deflection Load Case Deflection
~
~Total
5WL3

384EJ
WL3
t:=T4 fp

WL3
384EI
~
ta· bL c
:I a
384EI 1
~
~ L/2
I
L/2
t;;
V,
PL3

192EI
JU2
L
PL 3
t~ L/2  
l L ·1 48EJ ip b>a
1: r
~ ~
% 2 Pa 2 b 3
~ a J
b
b>a 3 EifJ
* L
~ b PL3
:1
c:S. .I
~a
a • 48EI 2
L
Wa 2 b

24EI
r:§?b •I
I a L ..
Wa 3
a
8EJ 3
t. a ~
t5'~
Ib
Wb4 Wab 3
+
8FJ 6FJ
1:
a
L J, :1 Pa 3
a
3EI 4
L a .Lb
p.l Wb3
+
3FJ
Wab2
3FJ
The maximum deflection of the beam carrying the uniform loading will oecur at the mid
5WL4
5 x (8B~~ax) L4 0.104 B.MaxL2 ·
0 ~ _e_
=
384EI 384EI EI
s: span 0.104B.MaxL2 L
u actual < 
 360 00 <
EI  360
where B.Max is the maximum bending moment due to unfactored imposed loads only .
. I> 37.4B.MaxL
00  E
Note: care must be taken to ensure that a consistent system of units is used. A similar
calculation can be carried out for any other span ratio.
coefficient
In Example 2.2 a 305 x 165 x 46 UB was found to be suitable for shear bending web
buckling and web bearing. Check the suitability with respect to deflection, assuming brittle
finishes to the underside of the beam.
Imposed load= 30 kN
~ Dead load = 60 kN
..L~  ~ ~ ~
50 kN 2.0 m 1 4.0 m 40 kN
~~~
6.0 m
50 kN
40 kN
Figure 2.30
Flexural Members 69
(50+30)2
Maximum bending moment due to unfactored imposed loads
2
80kNm
0.104B.MaxL 2 span
.•• 8actual ~ :::::
EI 360
Table 5
span = 6000 _
16.7 mm > 5actual
360 360
Section is adequate with respect to deflection
This check could have been carried out more accurately using the values given in Table
2 .1 of the text.
~ 30 kN
~
~+t
Figure 2.31
<5
actual
_ 5WL

3
PL 3a 4 a
384EJ + 48EJ L L
3[ ( )3]
= 8.3 + 5.7= 14 mm
In this case the approximate technique overestimates the deflection by less than 5%.
Provided that the estimated deflection is no more than 95% of the deflection limit, from
Table 5 the approximate answer should be adequate for design purposes and a more
accurate calculation is not required.
70 Design ofStructural Steelwork
2.15.1.1 A less rigorous alternative to evaluating the lateral torsional buckling moment
capacity than that given in Clause 4. 3. 7. 3 is given in Clause 4. 3. 7. 7. The method applies
to equal flanged rolled sections such as universal beams, universal columns and channel
sections.
In this method the buckling resistance moment Mb between lateral restraints is given by:
Mb =pbSx
where:
Ph is determined from Table 19 for the appropriate py value, A and X
A is the slenderness of the section taken as LF!r
LE is the effective length from Clauses 4.3.5 or 4.3.6 as before,
r is the radius of gyration about the member minor axis, i.e. ryy
X is the torsional index as before, i.e. from published section property tables or D/T
The slenderness value can be modified to allow for the shape of the bending moment
diagram between restraint points by using the correction factor 'n' from Table 20, or more
accurately using Clause 4. 3. 7. 6, as in the rigorous method demonstrated in section 2.4.4.1
of this text.
In Example 2.3 the buckling moment (Mh) was determined using the rigorous method to be
94.8 kNm. The conservative method in determiningMb is as follows:
30 kN 45 kN
~ Figure 2.32
In this case the bending moment diagram between the restraints at B and C IS a
combination of two cases given in Table 20.
Figure 2.33
It is evident from this method that the value of Mb is underestimated. In situations where
a section is proved to be inadequate by a small margin then use of the rigorous method may
prove worthwhile in identifying an acceptable economic section.
In Example 2.4 the Mb value using the Rigorous Method and grade 50 steel was found to be
197.8 kNm.
A 8 c D
Figure 2.34
72 Design ofStructural Steelwork
From section property tables:
Sxx = 1055 x 103 mm3 ryy = 38.5 mm X= 38.3
In this case the bending moment diagram between the restraints at B and C 1s a
combination of two cases given in Table 20.
c__J +
Figure 2.35
1.0+0.94
Average value of n ~ 0.97
2
The other checks such as shear, web buckling, web bearing and deflection are all the
same as before. The Conservative Method is only applicable to the evaluation of the lateral
torsional buckling moment Mb.
The Steel Construction Institute publishes member capacity tables to enable a rapid,
efficient check on a member subjected to axial load, bending moment or combined axial and
bending forces (Ref.: 11). The use of such tables to check shear, bending, web buckling and
web bearing is illustrated in the following example.
Flexural Members 73
Example 2.11 Beam with intermittent lateral restraint Use of safe load tables
It is proposed to use a 533 x 210 x 109 UB as a main roof beam spanning 12.0 m and
supporting three secondary beams at the midspan and quarter span points, as shown in
Figure 2.36.
(a) Using the design data and safe load tables given, check the suitability of the
section with respect to:
(i) shear,
(ii) bending,
(iii) web buckling and
(iv) web bearing.
(b) Check the suitability of the beam with respect to deflection assuming brittle
finishes.
Design data:
Assume:
(i) the stiff bearing length for web buckling and web bearing= 25 mm
(ii) lateral and torsional restraint to the compression flange is provided at the end
supports and positions of the secondary beams only.
Figure 2.36
Solution:
Design distributed load (1.4 X 5.0) + (1.6 X 6.0) = 16.6 kN/m
Design beam end reactions (1.4 X 10.0) + (1.6 X 20.0) =46kN
74 Design ofStructural Steelwork
~46kN
46kN 46kN 16.6kN/m
~.3. 0m  ·
168.6kN 118.8kN
Table 2 3
2.19.1 Shear
The safe loads in Table 2.2 give the value of Pv as determined from Clause 4.2.3
2.19. 2 Bending
The moment capacity Me, the section classification and the buckling moment of resistance
Mb for a range of effective lengths with various 'n ' values are given in Table 2. 3. The
values of Mb are calculated on the basis of Au= nuv.1 as defined in Clause 4.3. 7.5 and in
section 2.4.4.1 of this chapter.
Pw= C1+(blxC2)+(tpxC3)
where:
C 1 is the contribution from the beam,
C2 is the contribution from the stiff bearing length (excluding any additional flange
plates),
C3 is the contribution from any continuously welded flange plate if present,
b1 stiff bearing length,
tP thickness of additional flange plate
In this problem there are no additional flange plates present, therefore;
Pw= C1+(b1xC2)
76 Design ofStructural Steelwork
From Extracts Table 2.2 = 371 kN and C2 = 1.38 kN/mm
= 371 + (1.38 X 25) = 405.5 kN
>> 168.6 kN
Section is adequate in web buckling
Note: At the design stage the stiff bearing length required for web buckling and web
bearing is often not known, a conservative value of Pw and Pcrip can be evaluated
assuming this to be zero.
In locations such as the column positions at midspan, the web buckling and
web bearing capacities should be calculated on the basis of the load being
distributed in two directions as shown in Figure 2.27, and the continuous
buckling/bearing values in Table 2.2 used. In this example the support is the more
critical location.
2.19. 5 Deflection
The deflection is checked for the unfactored imposed load only as shown in Figure 2.38
20 kN 20 kN 20 kN 6.0 kN/m
66kN~66kN 12.0m
Figure 2.38
Two pedestrian walkways are required for maintenance purposes in the machinery hall of a
petrochemical plant. It is intended to support the walkways using a series of universal
beams at 4 .5 m centres as shown in Figure 2.39 The proposed surface of the walkways
comprises an open grid flooring system which is attached to the top flange of the beams by
proprietary brackets. Using the data given, design a suitable universal beam section
considering;
(i) shear,
(ii) bending,
(iii) web buckling,
(iv) web bearing and
(v) deflection.
Design data:
Selfweight of open grid flooring (including handrails) 0.1 kN/m
Maximum imposed service load allowed on walkway 3.1 kN/m
Elevation
Part plan
Figure 2.39
Dead load = 10 kN
Imposed load = 30 kN Dead load = 8 kN/m
Imposed load= 10 kN/m
A 2.0m B 4.0m
c
6.0m
62 kN
27.2 kN/m
2.0 m 4.0m
VA ~~~ Vc
By proportion:
Vertical reaction at A= VA=
Vertical reaction at C = Vc =
122.9 kN
102.3 kN
Shear Force Diagram Design shear force
122.9 kN
JllexuralAfe~nbers 79
(ii) Shear
Clause 4.2.3 Fv= 122.9 kN Pv = 0.6pyAv = 0.6pytD
(iii) Bending
Clause 4.2.5 60% Pv = 0.6 X 338.9 = 203.3 kN > 68.5 kN Low shear
720
Shape factor= v = 646 = 1.11 < 1.2 :. py Sxx. governs
Table 7 . Sectlon
(1) . C1assttcatwn
'fi . & Pr
= (275) 2 = 1.0
. 2252
average load factor "'"   = 1.52
148.0
Me = py Sxx S 1.52py Zxx
pySxx = (275 x 785 x 103 ) I 10 6 = 215.9 kNm
1.52py Zxx = (1.52 x 275 x 653 x 103) I 106 = 273 kNm
Me= 215.9 kNm > Mx Section is adequate
in bending
82 Design ofStructural Steelwork
23.0 kN 23.0 kN
t=t
28.1 kN 33.9 kN
28.1 kN
8 .9 kN
31 .9
X
33.9 kN
7
Design shear force = 33.9 kN
Design bending moment = 52.5 kNm
Coincident shear force = 24.1 kN
Note: The coincident shear force used is adjacent to the
position of maximum bending moment
sxx  1096
Shape f:actor v    = 1.15 <1.2
Zxx 950
.. Me= pySxx = (275 X 1096 X 10 3)/10 6 = 301.4 kNm
3
Pcrip = (21 .1 X 7.6 X 275)/10 = 44.1 kN
Pcrip > 33.9 kN Adequate in web
bearing
Clause 2.5 Assuming no brittle finishes 0 :::;; span = 8000 = 40 mm
200 200
Oactual
, ____
0.104B.MaxL_ 2
EI
Unfactored imposed load due to each walkway
= 3.} X 4.5 = 14.0 kN
14.0 kN 14.0 kN
1.75 kN
15.75 kN
3
b'actual
14x83 x108 [3x2 4 (2) ] +
48 X 205 X 103 X 21370 8 8
1 8 [
14 X 8· X 10
48 X 205 X 103 X 21370 8
()3]
3 X 1 _ 4 l_
8
= 3 6 II1ffi
.
3 Axially Loaded Members
Introduction
The design of axially loaded members considers any member where the applied loading
induces either axial tension or axial compression. Members subject to axial forces most
frequently occur in bracing systems, pinjointed trusses, lattice girders or suspension
systems, as shown in Figure 3 .1
Wind bracing
members
Suspension cables
Cablestayed bridge
Figure 3.1
Frequently, in structural frames, sections are subjected to combined axial and bending
effects which may be caused by eccentric connections, wind loading or rigidframe action.
In most cases in which UB and UC sections are used as columns in buildings, they are
subjected to combined axial and bending effects. The design of such members is discussed
and illustrated in Chapter 4.
The types of section used for axially loaded members range from rolled uniform beams,
columns and hollow sections to threaded bars, flat plates and wire ropes.
The following discussion relates primarily to pinjointed structures which comprise the
majority of structures with members subject to axial loads only.
86
Axially Loaded Members 87
3.2 Pinjointed Frames
The use of beams and plate girders as discussed in Chapters 2 and 6 does not always
provide the most economic or suitable structural solution when spanning large openings. In
buildings which have lightly loaded long span roofs, when large voids are required within
the depth of roof structures for services, when plate girders are impractical, or for
aesthetic/architectural reasons, the use of roof trusses or lattice girders may be more
appropriate.
Trusses are frequently used as secondary structural elements to distribute wind loading
to the foundations, as temporary bracing during construction and for torsional and lateral
stability.
Roof trusses and lattice girders are openweb flexural members which transmit the
effects of loads applied within their spans to support points by means of bending and shear.
In the case of beams, the bending and shear is transmitted by inducing bending moments
and shear forces in the crosssections of structural members. Trusses and lattice girders,
however, generally transfer their loads by inducing axial tensile or compressive forces in the
individual members. The form of a truss is most economic when the arrangement is such
that most members are in tension.
The magnitude and sense of these forces can be determined using standard methods of
analysis such as the method of sections, joint resolution, tension coefficients, graphical
techniques or the use of computer software.
The arrangement of the internal framing of a roof truss depends upon its span. Rafters
are norn1ally divided into equal panel lengths and ideally the loads are applied at the node
points by roof purlins. Purlin spacing is dependent on the form of roof cladding that is used
and is usually based on manufacturers' data sheets. In instances where the purlins do not
coincide with the node points, the main members (i.e. the rafters, or the top and bottom
booms of lattice girders) are also subjected to local bending which must be allowed for in
the design.
The internal structure of trusses should be such that, where possible, the long members
are ties (in tension), while the short members are struts (in compression). In long span
trusses the main ties are usually cambered to offset the visual sagging effects of the
deflection.
In very long span trusses, e.g. 60 metres, it is not usually possible to maintain a constant
slope in the rafter owing to problems such as additional heating requirements caused by the
very high ridge height. This problem can be overcome by changing the slope to provide a
mansardtype truss in which the slope near the end of the truss is very steep," while it is
shallower over the rest ofthe span.
A few examples of typical pitched roof trusses are illustrated in Figure 3.2. Lattice
girders are generally trusses with parallel top and bottom chords (known as booms) with
internal web bracing members. In long span construction they are very useful since their
relatively small span/depth ratio (typically 1110 to 1/14) gives them an advantage over
pitched rooftrusses.
88 Design ofStructural Steelwork
The two most common types are the PrattTruss (NGirder) and the·Warren Truss, both
of which are shown in Figure 3.3. The top chord often has a camber to assist drainage from
the roof.
tiJJ~Llz00j
~ Pratt or N type truss ~
Figure 3.3
As with roof trusses, the framing should be triangulated, considering the span and the
spacings ofthe applied loads. If purlins do not coincide with the panel points then secondary
Axially Loaded Members 89
bracing as shown in Figure 3.3 can be adopted as an alternative to designing for combined
axial and local bending effects (see Example 3.6).
Generally, the four main assumptions made when analysing trusses are:
In practice, the top and bottom chords are normally continuous and span several joints
rather than being a series of discontinuous, short members. Since truss members are usually
long and slender and do not support significant bending moments this assumption in the
analysis is acceptable.
The design of such members is carried out on the basis of combined axial and
approximate bending moment effects, as illustrated in Chapter 4, Example 4.4 and given in
Clause 4.8 ofBS 5950:Part 1.
In real trusses the members are connected at the joints using bolted or welded gusset
plates or end plates as shown in Figure 3 .4, rather than frictionless pins.
Provided that the settingout lines of the bolts or the centroidal axes of the members
intersect at the assumed joint locations, experience has shown that this idealisation is
acceptable.
Often the exact location of purlins relative to the joints on the top of the compression
chord/rafters is unknown at the design stage of a truss. In these circumstances, assuming
that the purlins do not coincide with the position of the joints, a local bending moment in
addition to the axial load is assumed in the truss members.
A nmnber of empirical design rules are given in Clause 4.10 for this situation. In
addition, Clause 4.10 indicates that secondary stresses induced by the inherent stiffuess of
the assumed pinjoints between the members, may be ignored provided that:
90 Design ofStructural Steelwork
(i) the slenderness of the chord members in the plane of the truss is g.ceater than 50 and
(ii) the slenderness of most of the web members is greater than 100.
+ The selfweight of the members may be neglected or assumed to act at the adjacent
nodes.
The following resume gives a brief summary of the mc.st common manual techniques
adopted to determine the forces induced in the members of statically determinate pinjointed
frames. There are numerous structural analysis books available which give comprehensive
detailed explanations of these techniques.
The method of sections involves the application of the three equations of static equilibrium
to twodimensional plane frames, i.e.
+vet "i..Fy =0 the sum of the vertical forces equals zero equation (i)
+ve~ "LFx =0 the sum of the horizontal forces equals zero equation (ii)
+vel "i.M =0 the sum of the moments of the forces taken
about anywhere in the plane of the frame
equals zero equation (iii)
The sign convention indicated in equations (i) (ii) and (iii) for positive directions has
been adopted in this text. The sign convention adopted to indicate ties and struts in frames is
....
as shown in Figure 3.5
.....
I I
Joint Strut  compression member Joint
• .... •
Tie  tension member
Figure 3.5
Axially Loaded Members 91
The method involves considering an imaginary section line which cuts the frame under
consideration into two parts A and B, as shown in Figure 3.8.
Since only three independent equations are available any section taken through a frame
must not include more than three members for which the internal force is unknown.
Consideration of the equilibrium of the resulting force system enables the magnitude and
sense of the forces in the cut members to be determined.
80 kN 80 kN 80 kN
Step 1: evaluate the support reactions. It is not necessary to know any information
regarding the frame members at this stage other than dimensions as shown in Figure 3. 7
since only externally applied loads and reactions are involved.
80 kN 80 kN 80 kN
STRUCTURE
~~ 80m
16.0 m
Figure 3.7
The moment equation can be considered about any point in the plane of the frame. Point
"A " is used since the line of action of both HA and VA passes through this point and
consequently these unknowns do not appear in the resulting equation.
Step 2: select a section through which the frame can be considered to be cut and using
the same three equations of equilibrium determine the magnitude and sense of the unknown
forces (i.e. the internal forces in the cut members).
I
80 kN 80 kN 80 kN
j j
,/section line
zero E ; F
~~~~?l~~~~I~?l~~~~~j
150 kN N / M 90 kN
:'
section line
Part A Part B
Figure 3.8
It is convenient to assume all unknown forces to be tensile and hence at the cut section
their direction and lines of action are considered to be pointing away from the joint (refer to
Figure 3.8). If the answer results in a negative force, this means that the assumption of a tie
was incorrect and the member is actually in compression, i.e. a strut.
The application of the equations of equilibrium to either part of the cut frame will enable
the forces X, Y and Z to be evaluated.
Note: the section considered must not cut through more than three members with unknown
internal forces.
Consider Part A
80 kN 80 kN 80 kN
~
Part A
50 kN
. 3 4
Note: sm {)=  = 0.6, cos {)=  = 0.8
5 5
+vet 'LFy = 0 150 (80 + 80 + 80) + FNFsinB = 0
FNF __.22_ = 90 = +150 kN
sin{) 0.6
Member NF is a TIE
These answers can be confirmed by considering Part B of the structure and applying the
equations as above.
Considering the same frame using joint resolution highlights the advantage of the method of
sections when only a few member forces are required.
In this teclmique, (which can be considered as a special case of the method of sections),
sections are taken which isolate each individual joint in tum in the frame, e.g.
80 kN 80 kN 80 kN
150 kN M L K 90 kN
Figure 3.10
In Figure 3.10 eight sections are shown, each of which isolates a joint in the structure as
indicated in Figures 3 .11 (a) and (b).
FLG F
.
JOINT G JOINT M JOINT F JOINT E
Figure 3.11 (b)
Since in each case the forces are coincident, the moment equation is of no value; hence
only two independent equations are available. It is necessary when considering the
equilibrium of each joint to do so in a sequence which ensures that there are no more than
two unknown member forces in the joint under consideration. This can be carried out until
all member forces in the structure have been determined.
Consider Joint A:
+vet I.Fy = 0 +90 Fw...cosfJ 0
FJH~
90
Fw... = + +150 kN
FJK 90 kN 0.6
+ve ~ "i.Fx = 0  Frn Fw...sjnfJ = 0
cosfJ = 0.6 Frn = (150 x 0.8) 120 kN
sinfJ = 0.8 Member JK is a TIE and member JH is a STRUT
Consider Joint K:
+vet I.Fy = 0 +FKH + 150sinB = 0
FKH = (150 X 0.6) 90kN
+ve ~ "i.Fx = 0  FKL + 150cosB = 0
Consider Joint H:
+ve t I.Fy = 0 +90  FHLcosfJ = 0
FHG ~~fJ"'tr•~ 120 kN 90
FHL = + = +150 kN
FHL 90 kN 0.6
+ve ~ "i.Fx = 0
120 FHG FHLsinfJ = 0
cosfJ= 0.6 =  (150 X 0.8)
FHG =  20 kN
sinfJ = 0.8 Member HL is a TIE and member HG is a STRUT
Consider Joint L:
150 kN
+vet I.Fy =0 +Fw + 150sinB = 0
FLG
= 90kN

Fw =(~50 X 0.6)
k{ +ve ~ "LFx =0  FLM + 150cosB+ 120 =0
FLM 120kN
FLM = (150 X 0.8) + 120 = +240 kN
Axially Loaded Members 95
sinO= 0.6
cosO= 0.8 Member LG is a STRUT and member LM is a TIE
Consider Joint G:
+vet L.Fy = 0 +90  FGMcosfJ = 0
FGF /1fJ~, 240 kN 90
FGM = +  = +150 kN
FGM 0.6
90 kN
+ve) L.Fx = 0  FGF 240 FGMSinfJ = 0
cosfJ = 0.6 FGF = (150 X 0.8) 240 = 360 kN
sin/3 = 0.8 Member GM is a TIE and member GF is a STRUT
Consider Joint M:
FMF 150 kN +vet L.Fy = O +FMF 150sin0 = 0
FMF = (150 X 0.6) = 90 kN
Consider Joint F:
FFE 360 kN
+vet L.Fy = 0 + 90 FFNcosfJ = 0
90
FFN = + = +150 kN
0.6
;;;t:kN
FFN I +ve) L.Fx = 0  FFE FFNsinfJ 360 = 0
FFE = (150 X 0.8) 360 = 480 kN
cos/3 = 0.6
sin/3 = 0.8 Member FN is a TIE and member FE is a STRUT
Consider Joint E:
80 kN +vet L.Fv = 0 FEN 80 =0
l = 80kN
r.
FEN
FED 480 kN +ve ) L.Fx = 0 FED 480 =0
FED = 480 kN
Member EN is a TIE and member ED is a STRUT
Graphical techniques have largely been superseded by the use of computer software for
all but the most basic trusses. In addition, tables of coefficients are available which enable
the rapid evaluation of member forces for a variety of structural forms, e.g. Timber
Designers 'Manual [Ref:20].
96 Design of Structural Steelwork
The method of tension coefficients is a tabular technique of carrying out joint resolution in
either two or three dimensions. It is ideally suited to the analysis of pinjointed spaceframes.
Consider an individual member from a pinjointed planeframe, e.g. member AB shown
in Figure 3 .12 with reference to a particular XY coordinate system.
If AB is a member of length LAB having a tensile force in it ofTAB, then the components
of this force in the X andY directions are TAB CosO and TAB SinO respectively.
If the coordinates of A and B are (XA, YA) and (XB, YB), then the component of TAB in
the xdirection is given by:
Ys 8
Ys
r~~~ X
0 XA Xs
Figure 3.12
T,
where tAB = and is known as the tension coefficient of the bar. Similarly, the
___1&.
LAB
component of TAB in the ydirection is given by :
( Y,  y )
ycomponent= TAB B
A = tAB(YB YA)
LAB
If at the joint A in the frame there are a number of bars, i.e. AB, AC ... AN, and external
loads XA and YA acting in the X and Y directions, then since the joint is in equilibrium the
sum of the components of the external and internal forces must equal zero in each of those
directions.
Expressing these conditions in terms of the components of each of the forces then gives:
Axially Loaded Members 97
A similar pair of equations can be developed for each joint in the frame giving (2 x No.
of joints) equations in total. In a triangulated frame the number of unknown forces (i.e.
members) is equal to [(2 x No. of joints) 3], hence there are three additional equations
which can be used to determine the reactions or check the values of the tension coefficients.
Once a tension coefficient (e.g. tAB) has been determined, the unknown member force is
given by the product
Note: A member which has ave tension coefficient is in compression and is therefore a
strut.
10 kN
B 20kN
k<      ,  
3.0m
Ax
4.0m
Ay Cy
Figure 3.13
The equilibrium equations are solved in terms of the 't' values and hence the member
forces and support reactions are evahmted and entered in the table as shown in Table 3 .1.
98 Design ofStructural Steelwork
Table 3.1
Consider joint B:
There are only two unknowns and two equations, hence:
Adding both equations
 4tAB + 3fBc + 20 0
 3tAB 3tBc 10 = 0
7fAB + 10 0 tAB 1.43
Joints A and C can be considered in a similar manner until all unknown values, including
reactions, have been determined. The reader should complete this solution.
In the case of a space frame, each joint has three coordinates and the forces have
components in the three orthogonal X, Y and Z directions. This leads to (3 x No. of joints)
equations which can be solved as above to determine the 't' values and subsequently the
member forces and support reactions.
The space frame shown in Figure 3 .14 has three pinned supports at A, B and C, all of
which lie on the same level as indicated. Member DE is horizontal and at a height of 10 m
above the plane of the supports. Determine the forces in the members when the frame
carries loads of 80 kN and 40 kN acting in a horizontal plane at joints E and D respectively
(see Table 3.2)
Axially Loaded Members 99
,5 m~_ 5
m,
I 5 m
I J
~~~ BOkN
10m
Figure 3.14
As indicated in Section 3 .1, the types of element used in the design of tension members is
numerous and varied; open and closed singlerolled sections being adopted for light trusses
and lattice girders, compound sections comprising either multiplerolled sections or welded
plates for heavy trusses with ropes and cables being used in suspension structures such as
bridges and roofs.
There are a number of potential problems that may arise from using light, slender
sections such as bars, flats, rolled angle and channel sections, e.g.
The introduction of sag rods as indicated in Figure 3.2 and the use of intermediate
packing in double angle or channel members will assist in minimizing the first two of these
problems.
In general, if the leg length of an angle tie is at least equal to member length , the member
60
will have sufficient stiffuess to prevent damage during transport.
where:
py is defined in Table 6 and
Ae is the effective area of the section as defined in Clauses 3. 3.3 or 4. 6. 2 to 4. 6. 4
100 Design ofStructural Steelwork
Table 3.2
B X DE + 4.0 10 + 40.0
y EC 4.0 15 60.0
z EB + 4.0 15 + 60.0
c X Support Reactions
y Support Component Reaction
z A Ax 0
D X Ay 0
y Az 0
z B Bx 40.0
E X By 20.0
y Bz 40.0
z c Cx 40.0
Clause 3.3.3
Experimental testing has indicated that the effective capacity of a member in tension is not
reduced by the presence of holes provided that the ratio:
net cross sectional area yield strength b . bl .
> y a smta e margm.
gross cross sectional area ultimate strength
This margin is reflected in the differing K., values adopted for different grades of steel, as
indicated in Table 3.3.
Table 3.3
where:
Us is the specified minimum ultimate tensile strength
Ys is the specified minimum yield strength
30mm ~~ 80 mm .I
Flat plate tie member Cross section
Figure 3.15
102 Design ofStructural Steelworlf
Ft g[ ·414+4
i I i I i
+4+1+++1
 Ft w~.....~4~+~~~~§ ~ t
j y j y j
Cross section
i
I.Sp .I.Sp . I. Sp .I. Sp .I
Figure 3.16
where:
sp is the staggered pitch
g is the gauge
t is the thickness
For example, consider the flat plate tie member shown in Figure 3.17.
30mm
~~~~
30mm Crosssection
50 50 50
20 mm diameter
black bolts
Path 1 is the most severe with the highest area to be deducted and therefore:
L
Singleangle Singlechannel Tsection
Effective Area . Ae = a 1 + ( 3a 1
3a1 +a2
)a 2
Figure3.18
where:
a1 is the net sectional area of the com1ected leg
az is the sectional area of the uncom1ected leg
Note: the area of a leg of an angle is defined in Clause 4. 6. 3. 2 as ' ... the product of the
thickness by the length from the outer corner minus half the thickness, and the area of the
leg of a Tsection is the product of the thickness by the depth minus the thickness of the
flange ', as shown in Figure 3.19.
104 Design ofStructural Steelwork
(d t)
Angle leg area
J
= (A  0.5t)t
T section leg area = (d t)t
Figure 3.19
r// ////1
Effective Area
Figure 3.20
where a1 and az are as before.
In double angle ties where x > aggregate thickness of the legs connected with solid
packing pieces or the slenderness of an individual angle > 180, each angle should be
regarded as a single angle and the effective area calculated as in Clause 4. 6. 3.1.
Clause 4.6.3.3
In symmetrical connections such as two angles backto back or members connected using
lug angles as shown in Figure 3 .21, the effective area should be calculated as for flat plates
using Clauses 3.3.2 to 3.4.
Axially Loaded Members 105
Consider the lattice girder in Example 3.1 and Figure 3.6, and check the suitability of the
sections shown in Table 3.4 for members NM and JK. The applied loading is assumed to be
the factored design loading and 16 mm diameter black bolts are to be used throughout.
Table 3.4
Clause 4. 6.1
Table 6
The two angles should be held apart by bolted packing pieces in at least two places
along their length, as shown in Figure 3 .22, using the same bolt size as the end connection.
Clause 4. 6.3.3
The outermost packing piece should be approximately (9 x smallest leg length) from the end
connection, i.e. (9 x 50) = 450 mm.
Although Clause 4. 6. 3. 2 specifically states that in double angles connected back to back
to one side of a gusset or section, the slenderness of the individual components should not
exceed 80; this is usually also applied to double angles connected to both sides of a gusset
or section, therefore:
t i t t t f ij
)
~
I !
450 775 775 775 775 450
.. I .. • I • .. I..
:I
• I .. 1 •
1: 4000 mm
Figure 3.22
The design of compression members is more complex than that of tension members and
encompasses the design of structural elements referred to as columns, stanchions or struts.
The term strut is usually used when referring to members in lattice/truss frameworks, while
the other two generally refer to vertical or inclined members supporting floors and/or roofs
in structural frames.
Axially Loaded Members 107
As with tension members, in many cases such members are subjected to both axial and
bending effects. This chapter deals primarily with those members on lattice/truss frame
works in which it is assumed that all members are subjected to concentric axial loading.
Column/stanchion design in which combined axial compression and bending are present is
discussed in Chapter 4.
The dominant mode of failure to be considered when designing struts is axial buckling.
Buckling failure is caused by secondary bending effects induced by factors such as:
The first two of these factors are the most significant and their effect is to introduce initial
curvature, secondary bending and consequently premature failure by buckling before the
stress in the material reaches the yield value. The stress at which failure will occur, known
as the compressive strength (pc), is influenced by several variables, e.g.
The effects of these variables are reflected in Tables 25, 26 and 27(a), (b), (c) and (d) in the
code, which ar2e used to determine the appropriate value of Pc for a particular
circumstance.
A practical and realistic assessment of the critical slenderness of a strut is the most
important criterion in determining the compressive strength.
Limiting values 'of slenderness are given in the code to reduce .the possibility of
premature failure in long struts, and to ensure an acceptable degree of robustness in a
member. These limits, which are given in Clause 4. 7.3.2 are shown in Table 3.5 of this
chapter.
Table 3.5
In addition to the values in Table 3.5, members in which the slenderness is greater than
180 should be checked for selfweight deflection. In cases where this deflection is greater
than (0.001 x length), then the secondary bending effects should be considered in the design.
The effective length is considered to be the actual length of the member between points of
restraint multiplied by a coefficient to allow for effects such as stiffening due to end
connections of the frame of which the member is a part. Appropriate values for the
coefficients are given in Table 24 of the code and illustrated in Figure 3.23(a) and (b).
In the case of angles, channels and T sections, secondary bending effects induced by end
connections can be ignored and pure axial loading assumed, provided that the slenderness
values are determined using Clauses 4. 7.1 0. 2 to 4. 7.1 0. 5 or Table 28 in the code. The use
of these Clauses and Table 28 is illustrated in Examples 3.5 and 3.6.
In the case of other crosssections the slenderness should be evaluated using effective
lengths as indicated in Figures 3.23(a) and (b). In addition, Appendix D ofthe code gives
the appropriate coefficients to be used when assessing the effective lengths for columns in
singlestorey buildings using simple construction; this is dealt with in Chapter 4.
Figure 23(a)
Axially Loaded Members 109
Sway Sway Sway and rotation
Figure 3.23(b)
The coefficients given for determining effective lengths are generally greater than those
predicted by mathematical theory; this is to allow for effects such as the inability in practice
to obtain full fixity.
The gross crosssectional area of a member is determined using the member profile and size
ignoring any holes required for fasteners, but allowing for any additional holes. When
considering the design of laced, battened struts or splices, the crosssection of the laces,
battens and splices should be ignored.
Where an element has been defined as slender using the criteria from Table 7, and it is
subject to compression, the design strength py should be modified by a factor which is given
in Table 8 of the code for various crosssections. This reduced value of py is then used when
extracting values of Pc (compressive strength) from Tables 27(a) to (d).
110 Design ofStructural Steelwork
3.9.2.3 Compressive Strength (Clause 4. 7.5)
When determining the value of Pc for a section, reference is first made to Table 25. If
additional flange plates have been welded to rolled I or Hsections, or if welded plate I or H
sections are used then further reference is made to Table 26. This is necessary since the
residual stresses induced by welding differ in both magnitude and distribution from those
caused during the rolling/cooling processes and effectively further reduce the stress at which
buckling may occur. Table 25 indicates the relevant Table 27 (a), (b), (c) or (d) to use in
determining the appropriate Pc value and relates to the type of crosssection, thickness of
elements and axis of buckling being considered.
Table 27 requires the py value and slenderness of the section to determine the tabulated
Pc value. When using Table 27 to determine the Pc value for sections fabricated by welded
plates, the py value should be reduced by 20 N/mm2 •
Compound struts comprising two components backtohack and connected symmetric
ally to both sides of a gusset are less stable about an axis through the plane of the
connection of the components than an equivalent solid section. The slenderness is therefore
checked using an additional ratio At, as given in Clauses 4. 7. 9 and 4. 7.13.
where:
An. is the ratio of the effective length to. the radius of gyration of a whole member about
the axis perpendicular to the plane of the connection,
A., is the ratio of the effective length of a main component to its minimum radius of
gyration.
Note: The main components should be connected at intervals such that the member is
divided into a minimum ofthree bays of approximately equal length.
+~ · 11·
9==1 =~~lr ~l! ~~~,=='
\; ·1: J'~ ~. ~~:=======Pi
~~
P = = = !
\; I 1
1
Consider the lattice girder in Example 3 .1 and Figure 3. 6 and check the suitability of the
sections shown in Table 3.5 for members FM and EF. The applied loading is assumed to be
factored design loading and 16 mm diameter blackbolts are to be used.
Axially Loaded Members 111
Table 3.5
b 75
t:= 1.0  7.5 < 8.5&
T 10
d 100
10.0 < 9.5&
t 10
b+d 175
 17.5 < 23&
T 10
Section is semicompact
Clause 4. 7.2 To determine effective length and slenderness use either Table 24 or Clause
4. 7.1 0. In structures such as the one considered here, it is normal to refer to Clause 4. 7.10
for discontinuous strut members and in the case of continuous members i.e. the compression
boom of a lattice girder, as an alternative to refer to Table 24.
lJVV slenderness ~
~
0.85Lwlrw
0.7Lwlrw + 15
b ·~· b
~ l.OL.Jraa
~ 0.7Laalraa+ 30
/ !
~ 0.85Lw1rbb
'..// !
!
v a ~ 0.7LbJrbb + 30
where:
L ••, Lbb and Lw are taken as the length 'L' between the intersections of the centroidal axes or
the settingout lines of the bolts.
Note: Clauses 4. 7.10.3 to 4. 7.10.4 give similar criteria for doubleangles struts, single
channel struts and single T section struts respectively.
Member EF: 2/150x75x 15 double angles, long leg connected to gusset plate.
Table 6 thickness < 16 mm :. Pv = 275 N/mm2
Table 7 b = 75 mm d = 150 mm T = 15 mm
b 75
&= 1.0 5.5 < 8.5&
T 15
d 150
= 10.0 < 9.5&
t 15
b+d 225
  15 < 23&
T 15
Section is semicompact
2: [(4vl'vyf +A: Y
2: 1.4A.,
where:
Lxx and Lyy are taken as the length 'L' between the intersections of the centroidal axes
or the setting out lines of the bolts.
Ac = 4v <50 and
rvv
Lvv is the distance between intermediate packing pieces
rvv is the radius of gyration for a single angle about the vv axis
Lxx = Lyy = 4000 mm rxx = 47.5 ryy = 31.0 mm rvv= 15.8 mm
For a 150x75xl5 double angle since A., ..s; 50 Lvv ..s; 50 x 15.8 790 mm
4000
Assume the member is divided into 6 bays :. Lvv = 667 mm
6
A
c
= 667
15.8
= 42.19
r
I
130 85 86 89
135 80 81 84
140 76 76 79 v
Extract from Table 27(c) BS 5950:Part 1
A rugby club requires a new stand and, while full uninterrupted views would be desirable,
the cost of a cantilever roof is prohibitive. An alternative solution is where a series of
castellated beams are simply supported on a series of rectangular hollow sections at the rear
and four Pratt trusses at the front, as shown in Figures 3.25 (a), (b) and (c). Using the
.design data given, check the suitability of the proposed section size for the members of the
trusses.
Design Data:
Characteristic dead load due to self weight of roof decking etc. 1.0 kN/m2
Characteristic imposed load 0.6 kN/m2
Top and bottom chords and diagonals 200xl20x8.0 RHS
As mentioned in Section 3 .2, when loads are applied to the chords of lattice girders
between the node points, a system of secondary bracing can be used to transfer the forces to
the node points and main girder members.
In this example each of the four main Pratt trusses supports a series of beams some of
which are located at the midspan points between the nodes in the top chord. The loads
applied to the chord by these beams can be transmitted through a secondary bracing system
using additional members, as shown in Figure 3.26. It is convenient when analysing this
structural arrangement to consider the truss as the superposition of two systems as shown.
Axially Loaded Members 115
Castellated beams
Pratt truss
r"'.:::;"" II II
II
l.Tr'"'...J II
rr 11
fl
11
II
II
[,;:] C,~J
}{ \\ II\\It
II
ll \\
II 11
u" \\
~ ::: \~ If \\
u ~ lj ~
16.0 m
II II II II II
II II II II II
~ 
  
ttl! I IT! I tt rr   11 X
20.0 m 20.0 m 20.0 m 20.0 m
f            80 .0 m       1
F/2 F F F F F F F F/2
~Sl$?=b±1
( 4 bays at 5.0 m each {
Actual loads system
F 2F 2F 2F F
ts::l~~'~l;?=l;?=l E
Primary load system
A ~Fe oFAcs
IFABS~
Secondary load system
Figure 3.26
The total force in any member is determined from the sum of the Primary and the
Secondary forces, for example;
Solution:
Area supported by each castellated beam ~ 2.5 x (5 + 16 + 3.5) 61.25m2
Design load/m 2 (1.4 x 1.0) + (1.6 x 0.6) 2.36 kN/m2
Total design load /beam 2.36 x 61.25 144.55 kN
Each castellated beam is simply supported with a 5. 0 m cantilever overhang
Axially Loaded Members 117
Figure 3.27
The support reaction F1 for the beam equals the magnitude of the point load applied to
the Pratt truss.
r
0 683 kN E 683 kN 0
Figure 3.28
4250
Clause 4. 7.3.2 87.5 s 180
rYY 48.6
Section is inadequate and must be increased in size such that Pc > 1002 kN
Bottom chord and diagonal members have maximum forces of 683 kN and 734 kN in
tension respectively.
A smaller section can be used, but for aesthetics and ease of detailing connections the
breadth 'b' of all members should be the same.
The reader should redesign the top chord member and design the secondary bracing
members CB and BD in a similar manner.
Design Loads:
F 1 = characteristic dead load Gk = 75 kN
F 1 = characteristic imposed load Qk =175 kN
F 2 = characteristic dead load Gk = 20 kN
F 2 = characteristic imposed load Qk = 75 kN
Axially Loaded Members 119
5.0 m
203x20 3x60 UC 
Figure 3.29
Solution:
Assuming the column to be effectively held in position at both ends, but not restrained in
direction at either end, Table 42 indicates an effective length equal to L.
Section Properties:
D = 209.6mm B = 205.8mm T 14.2 mm A = 7640 mm2
ryy = 52.0 mm rxx = 89.6 mm 9.4 mm dlt = 17.1 biT = 7.25
Columns which are assumed to be nominally pinned at their bases are provided with a slab
base comprising a single plate fillet welded to the end of the column and bolted to the
foundation with four holding down (H.D.) bolts. The base plate, welds and bolts must be of
adequate size, stiffuess and strength to transfer the axial compressive force and shear at the
support without exceeding the bearing strength of the bedding material and concrete base,
as shown in Figure 3.30.
r~~=t~~~:;;; Bedding
+B
I
'
I'
..,... 
+I '
._,...'
I
material
.· 1. Location cone ·I .
I " /
Anchor plates I
Figure 3.30
Clause 4.13. 2. 2 of BS 5950:Part 1 gives the following empirical formula for determin
ing the minimum thickness of a rectangular base plate supporting a concentrically loaded
column:
where:
a is the greater projection of the plate beyond the column as shown in Figure 3.31
b is the lesser projection of the plate beyond the column as shown in Figure 3.31
w is the pressure on the underside of the plate assuming a uniform distribution
pyp is the design strength of the plate from Clause 3.1.1 or Table 6, but not greater
than 270 N/mm2
Note: the value of 270 N/mm2 is a conversion of the value adopted in the previous
permissible stress design code for steelwork BS 449:Part 2: 1969 'The Use of
Structural Steel In Building'.
4B
I
•
I
I
I
4
0
I I
r' · r' =r a
I
Uniform pressure a>b ~b~
Figure 3.31
The code indicates that this formula may be used for I, H, channel, box or RHS sections.
The use of the equation, which assumes full twoway bending of the plate, can produce
plate thicknesses less than would be obtained at critical sections assuming simple cantilever
actions. An alternative approach considering a series of simple cantilever projections, which
is more appropriate for sections with a large depth/width ratio, e.g. Universal Beams, is the
Effective Area Method. This teclmique is comprehensively explained in Reference 14.
The dimensions of the plate must be sufficient to distribute the axial compressive load to
the foundations and to accommodate the holding down bolts.
The purpose of the welds is to transfer the shear force at the base and securely attach the
plate to the column. In most cases either 6 mm or 8 mm fillet welds run along the flanges
and for a short distance on either side of the web will be adequate.
The holding down bolts are generally cast within location cones in a concrete base and
fitted with an anchor plate to prevent pullout. The purpose of the location cone is to allow
for movement before final grouting and hence permit site adjustment during construction.
The diameter at the top of the location cones is usually at least 100 mm or 3 x bolt dia
meter. The recommended size ofH.D. bolts is M20 for light construction, M24 for bases up
to 50 mm thick increasing to M36 for heavier plates. Clearance holes in the base plates
should be 6 nun larger than the bolt diameter. The bolts are imbedded in the concrete base
to a length equal to approximately 16 to 18 x bolt diameter with a threaded length at least
equal to the bolt diameter plus 100 mm.
Bedding material can be either mortar, fine concrete or a proprietary, nonshrink grout.
In lightly loaded bases a gap of 25 mm to 50 mm is normally provided; this allows access
for grouting the H.D. bolt pockets and ensuring that the gap under the base plate is
completely filled. It is normal for the strength of the bedding material to be at least equal to
that of the concrete base. In BS 5950: Part 1 Clause 4.13.1, the bearing strength for
concrete foundations is given as 0 .4/~u, where .f~u is the characteristic concrete cube strength
at 28 days.
Design a suitable base plate for the axially loaded colunm in Example 3.6 in which the
design axial load :
F 986 kN Assume .feu = 40 N/mm2
25 X
10 ·96 (47.f OJ X 45.2 2 ) = 12.8 nun
270
Also ;;::: Flange thickness= 14.2 nun Assume 15 nun
H.D. Bolts
Assume M20's threaded length ~ 20 + 100 = 120nun
imbedded length ~ 16 x 20 320nun
thickness of bedding material 25nun
Total length of bolt = 120 + 15 + 25 + 320 480nun
4.1 Introduction
While many structural members have a single dominant effect, such as axial loading or
bending, there are numerous elements which are subjected to both types of loading at the
same time. The behaviour of such elements is dependent on the interaction characteristics of
the individual components of load. Generally, members resisting combined tension and
bending are less complex to design than those resisting combined compression and bending,
since the latter are more susceptible to associated buckling effects. The combined effects
can occur for several reasons such as eccentric loading or rigid frame action, as illustrated
in Figures 4.l(a) and (b) respectively.
..
,.,,,,,,,,,,,,,,,,,,,, ,,,,.,
!
/
Beam end reaction lrl
'1i I~
I I __\,
I
TF F
.J_
I
I F.e Rigid frame
i
I
8 ~Axial load
~ Shear force
Bending moment
At
Force system
(a) (b)
Figure 4.1
Consider a structural member subjected to concentric axial loading as shown in Figure 4.2.
The limiting value of applied axial load F can be determined by using the equation:
F
F = A.py :.   = 1.0
A.Py
Similarly, if the applied load is eccentric to the XX axis as shown in Figure 4.2(b):
= 1.0
123
124 Design ofStructural Steelwork
F
F.ey
1,1
Ill
II
1,1
111
II
+!
t t ~
ffi
I I
tH~
~ ~
(a) (b) (c)
Figure 4.2
If these limits are plotted on threedimensional orthogonal axes, then they represent the
members' capacity under each form ofloading acting singly. Figure 4.3(a) is a linear inter
action diagram. Any point located within the boundaries of the axes and the interaction
surface represents a combination of applied loading F, Mx and My for which
(1)
Interaction surface
1.0
My
Mcy
(a) (b)
Figure 4.3
Members Subject to Combined Axial and Flexural Loads 125
Equation ( 1) which represents a linear approximation of member behaviour is used in
BS 5950 in Clause 4.8.2. A more rigorous analysis allowing for plastic behaviour of plastic
and compact sections results in an interaction surface as shown in Figure 4.3(b). The
precise shape of the surface is dependent on the crosssection for which the diagram is
constructed. This nonlinear surface is represented by:
(2)
where:
Mrx the reduced moment capacity about the major axis due to axial loading
Mry the reduced moment capacity about the minor axis due to axial loading
(see Section 4.1.2 regarding reduced moment capacity)
z1 = 2.0 for I and H, solid and hollow circular sections
= 1.67 for solid and rectangular hollow sections
= 1. 0 for all other cases
z2 = 2.0 for solid and hollow circular sections
= 1.67 for solid and rectangular sections
= 1. 0 for all other cases
The values of Mrx and Mry are available from published Section Properties Tables. The
BS gives equation (2) as a more economic alternative to using equation (1) when designing
members subject to both axial tension and bending
Note: It is also necessary to check the resistance of such members to lateral torsional
buckling in accordance with Clause 4. 3 as shown in Chapter 2, Section 2.4.4, assuming the
value ofthe axial load to be zero.
In Chapter 2, Section 2.2 and Figure 2.3 the plastic stress distribution is shown for a cross
section subject to pure bending. When an axial load is applied at the same time, this stress
diagram should be amended. Consider an !section subjected to an eccentric compressive
! F.e( F.e~
~ F I
Figure 4.4
126 Design ofStructural Steelwork
The equivalent load on the crosssection comprises two components~ an axial load 'F
and a bending moment (F x e). The stress diagram can be considered to be the superposition
of two components as shown in Figure 4.5.
When the axial load is relatively low then sufficient material is contained within the web
to resist the pure axial effects as shown in Figure 4.5(a). There is a small reduction in the
material available to resist bending and hence the reduced bending stress diagram is as
shown in Figure 4.5(b). The stress diagram relating to combined axial and bending effects is
shown in Figure 4.5(c) indicating the displaced plastic neutral axis.
Plastic neutral
Compression
axis
zone
When the axial load is relatively high then in addition to the web material, some of the
flange material is required to resist the axial load as shown in Figure 4.6(a), and hence a
larger reduction in the bending moment capacity occurs as shown in Figure 4.6(b). As
before the plastic neutral axis is displaced as shown in Figure 4.6 (c).
Tension
t t ~;:;:;;;;~;;;;;;:;1zone
Plastic neutral
axis
Compression
zone
The published Section Property Tables provide a 'change formula' for each section to
enable a reduced plastic section modulus to be evaluated depending on the amount of
material required to resist the axial component of the applied loading.
Members Subject to Combined Axial and Flexural Loads 127
4.1.3 Combined Compression and Bending
The behaviour of members subjected to combined compression and bending is much more
complex than those with combined tension and bending. A comprehensive explanation of
this is beyond the scope ofthis text and can be found elsewhere (Refs:l2, 17). Essentially,
there are three possible modes of failure to consider:
1.0
Family of curves for
different values of A, and A,LT
Interaction
surfaces
Figure 4.7
Although similar, this interaction diagram differs from that shown in Figure 4.3 in that it
relates to slender members. The precise intercept on the F and Mx axes will depend on the
slenderness of the member. Clearly, short, stocky members will intercept the __!__ axis at
Agpy
a value of 1.0; as the slenderness A, increases, the intercept decreases. Similarly, in sections
which are fully restrained against lateral torsional buckling the intercept on the Mx ax1s
Mpx
(3)
which differs slightly from the equation given in Clause 4.8.2 for combined tension and
bending, or alternatively using the more exact method:
(4)
as before.
In addition an overall buckling check should be carried out in accordance with Clause
4.8.3.3.
A simplified, conservative approach to the design of members with combined axial
compression and biaxial bending is given in Clause 4. 8. 3. 3. 2 when considering overall
buckling. The following relationship should be satisfied:
(5)
(6)
where:
Max is the maximum buckling moment about the xx axis in the presence of axial load and
equals the lesser of:
M
(1£_)
~X or
ex ( 0.5F)
1+
pcx
May is the maximum buckling moment about the yy axis in the presence of axial load and
equals:
Members Subject to Combine,d Axial and Flexural Loads 129
[1£_)
Mcy [ pcy
1+ 0.5F
l
~y
where Pcx and P cy are the compression resistances about the maJor and mmor axes
respectively.
Condition 1:
When a beam is supported on a cap plate, as shown in Figure 4.8, the load should be
considered to be acting at the face of the column, or edge of packing if used, towards the
edge of the beam.
e e
h
Figure 4.8
Condition 2:
When a roof truss is supported using simple connections (see Figure 4.9), which cannot
develop significant moments, the eccentricity may be neglected and a concentric axial load
may be assumed at this point.
130 Design ofStru,ctural Steelwork
Bolted truss
support
Figure 4.9
Condition 3:
In all other cases, such as the beam connections shown in Figure 4.1 0, the eccentricity of
loading should be taken as 100 mm from the face of the column or at the centre of the length
of stiff bearing, whichever gives the greater eccentricity.
Note: The stiff length of bearing of a supporting element is defined in Clause 4. 5.1. 3 and
Figure 8. 0 of the BS. In most cases the designer does not know the precise details of say, a
seating angle, when designing a colunm and the assumption of 100 mm eccentricity will be
adopted.
Figure 4.10
When a crosssection is subject to combined axial and bending moments the classification
of the web should be checked for the web generally (see Table 7). For simplicity, the
classification may initially be conducted assuming pure axial load; if the result is either
plastic or compact, there is no advantage to be gained by carrying out a more complex,
accurate calculation. If the section is noncompact, then a more precise calculation should
be carried out (refer to Ref:l7).
An industrial unit comprises a series of braced rectangular frames as shown in Figure 4 .11 .
A travelling crane is supported on a runner beam attached to the underside, at the midspan
point of the rafters. Using the design data provided, check the suitability of a 203 x 133 x
25 UB for the colunms of a typical internal frame, using:
Members Subject to Combined Axial and Flexural Loads 131
Design Data:
Characteristic dead load due to the sheeting, purlins and services 0.5 kN/m2
Characteristic imposed load 0.75 kN/m2
Characteristic dead load due to side sheeting 0.3 kN/m2
Characteristic dead load due to crane 5.0kN
Characteristic imposed load 25.0 kN
Ignore wind loading
Typical
frame
\eating 1 . ~~atin.g/ ! l
cleat Crane load cleat I 5.0 m
Solution:
Area of roof supported by typical internal frame = (5.0 x 16.0) = 80m2
Design dead load due to roof sheeting etc. = 1.4 X (0.5 X 80) = 56 kN
Design imposed load on roof = 1.6 X (0.75 X 80) = 96 kN
Total distributed design load on rafter = (80 + 96) = 176 kN
Design load on rafter due to crane = (1.4 X 5.0) + (1.6 X 25) = 47 kN
Area of side sheeting supported by one column (5.0 x 5.0) = 25m2
176 kN
10.5 kN ~ 47 kN ! 10.5kN
Figure 4.12
132 Design of Structural Steelwork
Load imposed on column from end of rafter 0.5 X (176 + ~7) 111.5 kN
Total axial load on column (111.5 + 10.5) 122kN
Clause 4 7 6.(a).(3) An eccentricity equal to (DI2 + 100) should be assumed to evaluate the
bending moment since the precise details of the seating angles are not known at this stage.
In this case the web should be checked for "web generally" since both bending and axial
loads are present, if the section is plastic or compact when considering web to be in
compression throughout no precise calculation is required.
Table 7
d 798
< Web is plastic
t 84
Simplified approach
122 22.5
+ = 0.46 < 1.0
880 70.59
Members Su~ject to Combined Axial and Flexural Loads 133
An extract from member capacity tables published by the Steel Construction Institute is
given in Table 4.1 in which values are given for Mrx and Mry. In this case since Mry is equal
to zero the second term is not required.
Designation and FIPz Moment Capacity M,, M,. (kNm) and Reduced (
Axial Load Semi Moment Capacity Mrx, Mry (kNm) For Ratio of Axial
Capacity Compact Load to Axial Load Capacity FIP z
(Compact)
FIPz 0.0 0.1 0.2
     
Mcx 70.9 70.9 70.9
203xl33x25 Mcy 15.2 15.2 15.2
Pz= Agpy= 879 (1.00) Mrx 70.9 69.6 65.9
Mrv 15.2 15.2 15.2
mM_
_ _ mMY
x+  ::; 1.0
More Exact Method
Max M.y
mMX
sinceMy = zero ::; 1.0
Mx
(1_£_)
~X
Mcx (1 + 0.5F)
Max
::;
~y
Mcx = 70.95 kNm, Mb 27.9kNm
Pcy 208 kN Pcx (3200x241)/10 3 = 771 kN
( 1 122)
::; 70.95 771 = 55.34 kNm
( 1 + 0.5 X 122)
771
0.57x10.5
= 0.52 < 1.0
11.54
Section is adequate
Members Subject to Combined Axial and Flexural Loads 135
Note: In cases where either Mx or My approaches zero the more exact approach may be
more conservative (i.e. give a higher value) than the simplified approach. In such cases the
values satisfying the simplified approach may be used.
where:
Fe the compressive force due to axial load
Pc the compressive strength
Ag the gross crosssectional area
Mx the nominal moment about the major axis
My the nominal moment about the minor axis
Mbs the buckling resistance moment for simple columns determined as described for Mb in
Clause 4. 3. 7. 3 and 4. 3. 7. 4 but using an equivalent slenderness ratio Au of the
column given by Au= 5(Liryy).
py the design strength
136 Design ofStructural Steelwork
The floor plan and longitudinal crosssection of a twostorey, twobay braced st~elwork
frame is shown in Figures 4.13 (a) and (b). Beams B9 and BlOat the roof and first floor
level are small tie members, while all others are substantial members that are fixed to the
column flanges by web and seating cleats.
Check the suitability of a 203 x 203 x 52 UC for the column CB2 at section xx
indicated.
Design Data:
Characteristic dead load at roof level 4.0 kN/m2
Characteristic imposed load at roof level 1.5 kN/m2
Characteristic dead load at first floor level 6.0 kN/m2
Characteristic imposed load at first floor level 4.0 kN/m2
A 8 c
! 81
! 83
!
1 
89 I. 85 +86
I 6.0m
82 C82 84
2 
81 0 [ 87 +88
14.5m
3  81 83 _j
12.0 m 6.0 m I
~~
Figure 4.13(a)
I
Main beam Main beam
3.0 m
1st floor level I
l 5.0 m
c:::J
 Pinned bases
c:::J
Pinned base ~
c:::J
J
I 12.0 m I 6.0m I
Figure 4.13(b)
Members Subject to Combined Axial and Flexural Loads 137
Solution:
Section properties: 203 x 203 x 52 UC
D 206.2mm biT= 8.17 lxx = 5259 cm4 u = 0.848
B 204.3 mm d/t 20.4 !yy = 1778 cm4 X = 15.8
t 7.9mm 89.1 mm Sxx = 567 cm3 A = 6630mm2
T 12.5 mm 51.8 mm Zyy = 174 cm3 d 160.8 mm
The column CB2 supports four beams at each level. The slabs between gridlines A and
Bare oneway spanning and supported by beams B1 and B2. The slabs between gridlines
Band Care twoway spanning and supported by beams B3 to B8. Beams B2, B4 B5 and
B7 impose loads on column CB2 at both the roof and first floor levels. Since the column is
the same section throughout, the critical section for design will be section xx as indicated in
Figure 4.13 (b). At this location all loading originating from the roof beams is considered to
be axial, while the first floor beams will induce nominal moments due to the eccentricity of
the connections (see Clause 4. 7. 6) in addition to their axial effects.
i
3.0m
2.25 m
I
I I I I
12.0 m 2.25 m 1.5 m 2.25 m
Figure 4.14
Area of floor supported by beam B4= ( 6.0 X 3.0) + ( 1.5 X 6.0) 225
2.0 2.0
Axial load transmitted from the roofto column CB2 = (126 + 69.8 + 36 + 20.2)
= 252 kN
First Floor level
Table 2 Design load = (1.4 x 6.0) + (1.6 x 4.0) = 14.8 kN/m2
Load on beam B2 (31.5 x 14.8) = 466.2 kN End reaction 233.1 kN
Load on beam B4 = (17.4 x 14.8) = 258.1 kN End reaction 129.1 kN
Load on beam B5 = (9.0 x 14.8) 133.2 kN End reaction 66.6 kN
Design load beam B7 = (5.1 x 14.8) = 74.9 kN End reaction = 37.5 kN
Figure 4.15
Clause 4. 7. 7 Since the column is continuous and the same section above and below
section xx :
]upper =!_=OJ flower = !_ = 0.2 Jul Lu = 0.33 = 1.65 > 1.5
Lupper 3 Llower 5 JL I 4 0.2
The net nominal moments applied at the first floor level should be divided in proportion
to the stiffuesses of each length of column.
Members Subject to Combined Axial and Flexural Loads 139
The load system on the crosssection of the column is equal to the superposition of three
components F, Mx and My as shown in Figure 4.16
+, +,
[1", II l~i
1
[1", II
F =720 kN Mx =7.97 kN My= 1.14
kf't11,l iiII II
II
I iIIi i Ill i
I II I
ttJ
+ +
t!]
Axial
IJ]
Net moment
about xx axis
Ill
II
lol
111
II
~
Net moment
about yy axis
Figure 4.16
Section classification :
Table 7 &=1.0 biT = 8.17 < 8.5& Flanges are plastic
d/t = 20.4
Since the web is subject to combined axial and bending effects use criterion for the web
generally. However, if a check assuming pure axial loading results in a plastic or compact
section then precise calculation is not required.
dlt < 39& Web is plastic
Clause 4. 7. 7
L 0.5 X 5000
Au= 0.5 = 48.3
ry 51.8
~ T

245

265

275

\
   
40
45
238
227
254
242
262
250
s
50 217 231 238
55 206 219 ??f.. J
Clause 4. 7.2.(a)
Assuming that all the beams provide directional restraint to the column for buckling about
both axes, the effective lengths are:
Section is adequate
Members Subject to Combined Axial and Flexural Loads 141
Consider the lattice girder in Example 3.6, in which a series of castellated beams are
supported at the midspan points between the nodes in the top chord, as shown in Figure
4.17.
t=SJ~LZi
~
~
4 bays at 5.0 m each .J
~
Figure 4.17
An alternative to using secondary bracing is to design the top chord members to resist a
combined axial load and bending moment. In BS 5950: Part 1, Clause 4.10 simplified
empirical design rules are given for such members in lattice frames and trusses. The axial
loads are determined assuming that the connections are pinned (Clause 4.JO(a)), e.g. the
primary axial loads in Example 3.6. The bending moment can be approximated by
assuming a 'fixed end moment value' if the load position between the nodes is known, or
using WL/6 as indicated in Clause 4.JO(c).
The effective length of members can be determined taking into account the fixity of the
connections and the rigidity of adjacent members.
In this example the axial loads in the lattice girder and the bending moment diagram for
the top chord member AD are shown in Figure 4.18.
lw
56.8 kNm 56.8 kNm
At'>.
~
C QD 1 WL/4
Figure 4.18
Member AD:
Design shear force (0.5 X 90.81) = 45.4 kN
142 Design ofStructural Steelwork
Section properties :
A = 9550 mm2 dlt = 27.0 biT 17.0 rxx= 11.2cm ryy= 8.14cm
Zxx= 796 cm3 Sxx = 964 cm3 t 10mm
Table 25 Use Table 27(a) for buckling about either the xx or yy axis. In this case, the
critical slenderness is A.xx = 37.9
Simplified approach
Clause 4.2.5
RHS members have two vertical webs and it is unlikely that the applied shear force will
exceed (0.6 x shear capacity) and the moment capacity can normally be taken as :
If Sxx is greater than 1.2Zxx the 1.2 can be replaced by the average load factor.
Sxx = 964 cm3 1.2Zxx = 1.2 x 796 = 955 cm3
.'. Sxx > 1.2Zxx
908 56.8
+ 0.56 < 1.0
2626 265.4
Alternatively
Clause 4.8.3.2(b) More Exact Check
My= zero
From published tables with = 908 = 0.35
2626
Clause 4. 8. 2 Z1 = 5/3 Mrx. = 226kNm
Mx)zl =
(Mrx. (113.6)i = 0.32 < 1.0
226
__x
mMY S 1.0
mM +·
Max May
(1 908)
= 265.4 2502 143.1 kNm
( } + 0.5 X 908)
2502
1.0 X 113.6
0.79 < 1.0
143.1
Section is suitable
5. Connections
5.1 Introduction
Traditionally, the consulting engineer has been responsible for the design and detailing of
structural frames and individual members, while in many instances the fabricator has been
responsible for the design of connections and consideration of local effects. Codes of
Practice tend to give detailed specific advice relating to members and relatively little
guidance on connection design. This has resulted in a wide variety of acceptable methods of
design and details to transfer shear, axial and bending forces from one structural member to
another.
Current techniques include the use of black bolts, highstrength frictiongrip (H.S.F.G.)
bolts, fillet welds, butt welds and, more recently, the use of flowdrill techniques for rolled
hollow sections. In addition there are numerous proprietary types of fastener available.
Since fabrication and erection costs are a significant proportion of the overall cost of a
steel framework, the specification and detailing of connections is also an important element
in the design process.
The basis of the design of connections must reflect the identified load paths throughout a
framework, assuming a realistic distribution of internal forces and must have regard to local
effects on flanges and webs. If necessary, localised stiffening must be provided to assist
load transfer.
All buildings behave as complex threedimensional systems exhibiting interaction
between principal elements such as beams, columns, roof and wall cladding, floors and
connections. BS 5950:Part 1, Clause 2.1.2 specifies four methods of design which may be
used in the design of steel frames:
145
146 Design of Structural Steelwork
CONNECTIONS
Simple design
~
HSFY
Black  bolts Fillet welds
~
H.S.F.G. bolts
Figure 5.1
The design of connections requires analysis to determine the magnitude and nature of the
forces which are to be transmitted between members. In both bolted and welded connections
this generally requires the evaluation of a resultant shear force and, in the case of moment
connections, may include combined tension and shear forces.
These are most frequently used in pinjointed frames and braced structures in which lateral
stability is provided by diagonal bracing or other alternative structural elements. Typical
examples of instances in the use of simple connections such as in single or multistorey
braced frames, or the flange cover plates in beam splices are shown in Figure 5.2.
In each case, the shear force to be transmitted is shared by the number of bolts or area of
weld used and details such as end/edge distances and fastener spacings are specified to
satisfy the code requirements.
Connections 147
Figure 5.2
These are used in locations where, in addition to shear and axial forces, moment forces must
be transferred between members to ensure continuity of the structure. Typical examples of
this occur in unbraced single or multistorey frames , support brackets with the moment
either in the plane of or perpendicular to the plane of the connection and web cover plates in
beam splices, as shown in Figures 5.3 (a) (b) and (c) .
148 Design ofStructural Steelwork
~
Wind
loading
Unbraced pitched roof
Beam web splice portal frame
Figure 5.3{a)
Beam/column
connection
Wind
loading
Bracket connections with the moment perpendicular to the plane of the connection
e
~P/bracket
Pe x z 1
no. of bolts
2>2
n=l
I
'
0 I 0
:~·:
cpP!n P!ncp
cpP!n P!ncp
+
cpP!n P!ncp
Vertical shear due to P Rotational shear
n = number of bolts dueto(Pxe)
Figure 5.4
Similarly for a web splice, e.g. as shown in Figures 5.5(a) and 5.5(b).
150 Design ofStructural Steelwork
t' P..e
. p
cpP/ n
cp PI
cpP/n
Direct Rotational
shear shear
Individual bolt shear loads
Figure 5.5(b)
~p
y
I'' I
I
I
I
,.)._.. X
I
X
I
'
:
I
+
I' :I
iy
Total weld area =A
Figure 5.6
Connections 151
P Pe
Total vertical component Fv +CosO
Area ofweld !polar
e p
a

 Assumed line
of rotation
jo ~
~~3
jo ~
jo ~
jo ~
n
Pey3
F3= 3
NLY~
m=l
where F 3 is the maximum tensile load in the bohs which must be designed for the combined
the combined effects of the vertical shear and horizontal tension.
e P
Figure 5.8
The diinensions and strength characteristics of bolts commonly used in the U.K. are
specified in BS 3692 (precision hexagon bolts), BS 4190 (black hexagon bolts) and BS
4395 (HighStrengthFrictionGrip bolts). Washer details are specified in BS 4320. BS
4190 specifies two strength grades: Grade 4.6 which is mild steel (yield stress = 235
N/mm2) and Grade 8.8, which is highstrengthsteel (yield stress= 627 N/mm2).
The most commonly used bolt diameters are: 16, 20, 24 and 30 mm; 22 mm and 27 mm
diameter are also available, but are not preferred.
The usual method of forming site connections is to use bolts in clearance holes which are
2 nun larger than the bolt diameter for bolts less than or equal to 22 mm dia., and 3 nun
larger for bolts of greater diameter. Such bolts are untensioned and normally referred to as
Black bolts. In circumstances where slip is not permissible, such as when full continuity is
assumed (e.g. rigiddesign), vibration, impact or fatigue is likely, or connections are subject
to stress reversal (other than that due to wind loading), High Strength Friction Grip bolts
should be used. Precision hexagon bolts in close tolerance holes have been superseded by
H.S.F.G. bolts and are rarely used; they will not be considered further.
Black bolts transfer shear at the connection by bolt shear at the interface and bearing on the
bolts and plates as shown in Figure 5.9
Figure 5.9
BS 5950 specifies minimum and maximum distances forthe spacing of bolts, in addition to
end and edge distances from the centreline of the holes to the plate edges as indicated. The
requirement for minimum spacing is to ensure that local crushing in the wake of a bolt does
154 Design ofStructural Steelwork
not affect any adjacent bolts. The maximum spacing requirement is to ensure that the
section of plate between bolts does not buckle when it is in compression.
The requirement for minimum end and edge distances is to ensure that no end or edge
splitting or tearing occurs and that a smooth flow of stresses is possible. Lifting of the edges
between the bolts is prevented by specifying a maximum edge distance.
The values specified in Clause 6.2 and Table 31 are illustrated in Figure 5.10.
P. =P.As
where:
Ps is given in Table 32 as 160 N/mm2 for Grade 4.6 bolts and 375 N/mm2 for Grade 8.8
bolts
As is the crosssection resisting shear, normally this is based on the root of the thread
(tensile area At). lfthe shear surface coincides with the full bolt shank then the shank
area based on the nominal bolt diameter can be used.
~2.50
~ 14t
~200+16t
Figure 5.10
where:
t is the thickness of the plate,
d is the nominal bolt diameter, and
Pbb is given in Table 32 as 460 N/mm2 for Grade 4.6 bolts and 1035 N/mm2 for Grade
8.8 bolts.
Connections 155
where:
d is the nominal diameter of the bolt,
is the thickness of the plate,
e is the end distance,
Pbs is given in Table 33 as 460 N/mm2 for Grade 43 steel.
In the case of Grade 4.6 bolts, equation (i) will always be the lesser value provided that
the end distance 'e' is greater than twice the bolt diameter. When Grade 8.8 bolts are used
equation (ii) will always be the lesser value.
where:
At is the tensile area ofthe bolt (based on the root thread diameter)
p1 is given in Table 32 as 195 N/mm2 for Grade 4.6 bolts and 450 N/mm2 for Grade 8.8
bolts
5.2.1. 6 Bolts Subject to Combined Shear and Tension (Clause 6.3. 6.3)
When black bolts are subject to both shear and tension simultaneously then the following
relationship should be satisfied:
F: +F; 14
$.
P., ~
where:
H.S.F.G. bolts are manufactured from high strength steel so that they can be tightened to
give a high shank tension. The shear force at the connection is considered to be transmitted
by friction between the endplate and column flange plate, as shown in Figures 5.11(a) and
(b).
Figure 5.ll{a)
Shank tension
Shear force
Friction at interface j
j
Figure 5.ll{b)
The bolts must be used with hardened steel washers to prevent damage to the connected
parts. The surfaces in contact must be free of mill scale, rust, paint grease etc., since this
would reduce the coefficient of friction (slip factor) between the surfaces.
It is essential to ensure that bolts are tightened up to the required tension, otherwise slip
will occur at service loads and the joint will behave as an ordinary bolted joint. There are
several techniques which are used to achieve the correct shank tension; these are now
described.
A power or handoperated tool which is used to induce a specified torque to the nut.
These have projections which squash down as the bolt is tightened. A feeler gauge is used to
measure when the gap has reached the required size.
Connections 157
5.2.2.3 Partturning
The nut is tightened up and then forced a further half to threequarters of a turn, depending
on the bolt length and diameter
H.S.F.G. bolts are generally used in clearance holes. The clearances are the same as for
ordinary bolts. The design of H.S.F.G. bolts when used in shear, tension and combined
shear and tension is set out in Clause 6.4 of BS 5950:Part 1 and is illustrated below. The
possibility of bearing failure must also be considered.
where:
P0 is the minimum shank tension as specified in BS 4604
(this is the proofload given in published section tables)
K8 = I. 0 for bolts in clearance holes
j.l = slipfactor, for general grade bolts and untreated surfaces j.l = 0.45
etpbg
P,g =dtpbg :::;; 3
where:
d = nominal diameter of the bolt,
=thickness of the connected ply,
e = end distance,
pbg =bearing strength from Table 34.0
where P0 is as before
158 Design of Structural Steelwork
5. 2. 2. 7 Combined Shear and Tension (Clause 6.4.5)
When H.S.F.G. bolts are subjected to an external tensile force, the clamping action, and
hence the friction force available to resist shear, is reduced. Three conditions must be
satisfied:
(i) Fs <PsL
<Pbg
(ii) Ft <Pt
Fs ~
(iii) + 0.8 s 1.0
PsL Pt
The design of simple bolted connections is illustrated in Example 5.1 and Example 5.2.
A lap joint is shown in Figure 5.12 in which a single Grade 4. 6 16 mm diameter black bolt
is used. There is one shear interface and it is assumed that this passes through the threaded
portion of the bolt.
(a) Check the minimum and maximum edge and end distances.
(b) Determine the shear capacity of the connection with respect to:
(i) bolt shear,
(ii) bolt bearing,
(iii) plate bearing, and
(iv) plate tension capacity.
20 kN
* 25mm 25mm
I +·~·'13 :::
I''
* It is desirable to adopt the minimum edge distance +5 mm to
accommodate any enlargement which may be necessary on site.
Figure 5.12
A lap joint similar to Example 5.1 is shown in Figure 5.13; in this case there are two shear
interfaces and four 20 mm diameter black bolts. The outer plates are 8 mm thick, while the
inner plate is 12 mm thick. Determine (a) and (b) as in the previous example.
75 kN Bmm
12 mm
75 kN 8mm
45 mm 50 mm 5
Figure 5.13
The use of Grade 8. 8 bolts, which have more than twice the shear capacity of Grade 4. 6
bolts is more economic, particularly since the installation costs are the same.
The most common processes of welding used in connections are methods of fo.sion (arc)
welding. There are several techniques of fusion welding which are adopted, both manual
and automatic/semiautomatic, such as manual metal arc (MMA) and metal inert gas
(MIG). The MMA process is usually used for short runs in workshops or on site and the
MIG process for short runs or long runs in the workshop. The two most widely used types
of weld are fillet and butt welds.
160 Design of Structural Steelwork
Fillet welds, as illustrated in Figure 5.14, transmit forces by shear through the throat
thickness.
The design strength (pw) is given in BS 5950:Part 1, Table 36, (in the case of Grade 43
steel Pw = 215 N/rnm2), and the effective throat thickness as defined in Clause 6.6.5.3 is
normally taken as 0. 7 times the effective leg length. It is assumed that when using this value
the angle between the fusion faces lies between 60° and 90°. When the fusion faces are
inclined at an angle between 91° and 120° then the 0.7 coefficient should be modified as
indicated in Figure 5 .15.
Double fillet weld Single fillet weld Lap splice Fillet weld detail
Figure 5.14
Figure 5.15
In situations when 120° s a s 60° poor access to the acute fillet weld and a small
throat thickness on the obtuse fillet weld can create problems. In these situations a different
type of weld, e.g. a singlesided butt weld, may be more appropriate. BS 5950 indicates in
Clause 6. 6. 5. 4 that the strength of such acute and obtuse fillet welds should be demon
strated by testing.
Where possible a run of fillet weld should be returned around comers for a distance of
not less than twice the leg length. If this is not possible, then the length of weld considered
Connections 161
effective for strength purposes should be taken as the overall length less one leg length for
each length which does not continue round a comer, (Clause 6. 6. 5. 2).
In many connections welds are subject to a complex stress condition induced by multi
directional loading. The strengths of transverse fillet welds and longitudinal fillet welds
differ, end fillet welds being the stronger. In addition, in side fillet welds, large longitudinal
forces are concentrated locally on the member crosssection.
A significant variation in tensile stress occurs across the width of the tensile members
when the lateral spacing between weld runs is considerably larger than the length. These
effects are limited in BS 5950 by limiting the spacing Twas indicated in Figure 5.16. In
Clause 6. 6. 5. 5 the code also permits the vectorial summation of stresses to determine the
stress for which fillet welds should be designed; this is illustrated in later calculations.
Effective length
Figure 5.17
The strength of a full penetration butt weld is defined in Clause 6. 6. 6.1 and is generally
taken as the capacity of the weaker element joined, provided that appropriate electrodes and
consumables are used. In most situations when designing structural frames, full penetration
162 Design ofStructural Steelwork
welds are unnecessary and the additional costs involved in preparing Slich connections is
rarely justified unless dynamic or fatigue loadings are being considered. Butt welds are used
extensively in the fabrication of marine structures such as ships, submarines, offshore oil
installations and in pipework.
The design of simple fillet weld connections is illustrated in Example 5.3 and Example 5.4.
75kN
Throat area
Leg length 's
Figure 5.18
Clause 6.6.5
Table 36 Design strength of fillet weld Pw = 215 N/mm2
Clause 6.6.5.3
Strength of fillet weld /mm length = Pw x throat area
throat size 'a' = 0.7 x leg length's'
a = 0 ~ 7 x 6.0 = 4.2 mm
:. Strength = (215 x 4.2x 1.0) I 103
= 0.903 kN
Clause 6. 6.3.2 Effective length of weld required, L = 75/0.903 = 83 mm
A 65 x 50 x 8 angle tie is connected by the long leg to a gusset plate and is required to
transmit characteristic loads as indicated in Figure 5 .19. Design a suitable fillet weld.
Crosssection
Figure 5.19
~"'""'"'"'"'"'''"'~
65 mm :~ : :: Jt~~~::::~+~ p
"''''''''''''''''''''''''''~'~''''''''''~
y
Figure 5.20
The force is assumed to be transmitted through the centroid of the section Since the force
is applied nearer to weld Y than weld X, weld Y will transmit a proportionately larger share
of the total force.
Clause 6.6.5
Table 36 Design strength of fillet weld Pw = 215 Nlmm2
Clause 6.6.5.3 Assuming a 6 mm fillet weld throat a =0.7x6.0 4.2mm
There are three types of beam end connections which are commonly used in the fabrication
of steelwork:
+ All three types of connection are capable of transmitting at least 75% of the shear
capacity of the beam being connected, depending upon the depth of the plates
and/or the number of vertical rows ofbolts used.
+ Fin plates are the most suitable for the connection of beams which are eccentric to
columns or connections which are skewed.
+ End plates are the most suitable when connecting to column webs.
+ Fabrication and treatment do not present any significant problems for any of the
three types of connection.
It is important in simple design to detail beam end connections which will permit end
rotation of the connecting beam, thus allowing it to displace in a simply supported profile
while still maintaining the integrity of the shear capacity. This rotational capacity 1s
provided by the slip of the bolts and the deformation of the connection component parts.
Typical angle web cleats comprise two 90 x 90 x 10 angle sections bolted or welded to the
web of the beam to be supported, as shown in Figure 5 .21.
Figure 5.21
Connections 165
The 10 mm projection of the web cleats beyond the end of the beam is to ensure that
when the beam rotates, the bottom flange does not bear on the supporting member. The
positioning of the cleats near the top of the beam provides directional restraint to the
compression flange. When this positioning is used in addition to a length of cleat equal to
approximately 60% of the beam depth, the end of the beam can be assumed to have
torsional restraint.
This detail enables an effective length of compression flange of l.OL to be used when
designing a beam which is not fully restrained.
0 0
0 0
0 0
0 0
Figure 5.22
5. 4. 3 Fin Plates
Figure 5.23
166 Design ofStructural Steelwork
Experimental data has indicated that torsions induced by the plate being connected on
only one side of the web of a beam are negligible and may be ignored.
Recommendations for the detailing of fin plates as given in Joints in Simple
Construction, Vo1.2 (Ref: 14) are :
5.5 Example 5.5 Web cleat, end plate and tin plate connections
A braced rectangular frame in which simple connections are assumed between the columns
and rafter beam is shown in Figure 5.24. Using the data provided, design a suitable
connection considering;
Figure 5.24
The principle difference between 'simple design' and 'rigid design' of structural frames
occurs in the design of the connections between the elements. In the former the connections
are assumed to transmit direct and shear forces; in the latter it is necessary to transmit
moments in addition to these forces. The moments can be considered to be either:
In both cases the connections are generally designed using either H.S.F.G. bolts or
welding. There are a number of approaches to designing connections in moment resisting
frames. In most connections the problem is to identify the distribution of forces, moments
and stresses in the component parts. Other factors which must be considered are the overall
stiffuess of the connection, and the practical aspects of fabrication, erection and inspection.
In rigid frame design each component of the frame can be designed individually to
sustain the bending moments, shear forces and axial loads; the connections must then be
designed to transfer these forces. In many instances the connections occur where members
change direction, such as at the eaves and ridge of pitched roof portal frames. Such frames
are usually transported 'piecesmall' from the fabrication shop to the site and provision
must be made for sitejoints. High Strength Friction Grip bolts are ideally suited for this
purpose.
In welded connections, such as the knee joint shown in Figure 5.25, because the
compressive forces x andy are not collinear, an induced compressive force Z exists to
maintain the equilibrium of the forces. This force Z acts across the web plate at the comer
and in order to obviate the likelihood of the web plate buckling, comer stiffeners are
required as shown, to carry the force Z.
Typical site connections using H.S.F.G. bolts are shown in Figure 5.26. The cap plate
transmits the force in the tension flange by means of the H.S.F.G bolts while the force in the
compression flange is transmitted in direct bearing. Where the depth of the connection in
Figure 26(a) leads to forces of too large a magnitude to be transmitted reasonably, the depth
of the connection can be increased by the introduction. of a haunch, such as is shown in
Figure 26(b).
Cap plate
 _L 
End plate
Stiffener
plate each
side of web
(a) (b)
Figure 5.26
The uniform rectangular portal frame shown in Figure 5.28 is subjected to loading which
induces moments and shear forces at the knee joint as given in the data below. Using this
data, determine a suitable size ofH.S.F.G. bolt for the connection between the column and
the roof beam.
Connections 169
Figure 5.28
r S.F.G bolts (
/illetweld
Tensile force
Flange bolts
transmit tensile force
due to moment
Design Data:
Colunm I Beam Section 610 X 305 X 149 UB
Ultimate Design Moment at the knee joint 540 kNm
Ultimate Design Shear Force at the knee joint 380kN
An industrial frame building supports a light electric overhead travelling crane on brackets
bolted to the main columns as shown in Figure 5.30 below. Using the design data given
determine a suitable size ofH.S.F.G. bolt to connect the brackets to the columns.
a
Crane gantry girder
3 at 90 mm I
Figure 5.30
L
Assuming that the bracket rotates at the level of the bottom bolts {this is conservative)
then the maximum tensile force induced in the bolts = Tmax =
where:
n = number of vertical columns of bolts
e = eccentricity of applied loads
P = applied load
Ymax =distance from the centre of rotation ofthe bracket to the most distant bolt
= [ 27.5
+ ( 0.8 X
5~~~8 )] = 0.71 ~ 1.4
71.3
M20 H.S.F.G. bolts are adequate
Connections 171
5.7 Splices
Fabrication and/or transportation constraints sometimes dictate that beams or columns are
delivered on site in separate sections which require spliced connections during the erection
stage. In such circumstances it is necessary for the splice to transmit all of the forces, i.e.
bending moment, shear and axial forces, which exist at the location of the connection.
Spliced connections must provide adequate stiffness and continuity; this is particularly
important at locations which are not adjacent to lateral restraints. The tendency for
compressive loads to induce lateral instability must be considered and can be accommodated
by either:
(i) providing flange plates of similar dimensions to the flanges of the members being
connected, or
(ii) satisfying the requirements given in Appendix C: Clause C. 3 of BS 5950:Part: I
relating to strut action.
The design of splices should satisfy the recommendations given on Clause 6.1. 7 of the
code.
Beam splices can be either bolted or welded. Bolted splices using flange and web cover
plates as shown in Figure 5.31 (a) and (b) provide more rotational rigidity than welded
moment endplate splices and are the type considered in this text.
Figure 5.31(a)
Figure 5.30(b)
172 Design ofStructural Steelwork
Since in most cases, continuity of the beam section has been assumed in the analysis,
and joint rotation caused by bolt slip is generally unacceptable, it is recommended that high
strengthfrictiongrip bolts are used. Usual practice is to adopt a symmetrical arrangement
of bolts in each set of cover plates.
Spacing and edge distances as specified in Clause 6. 2 and 6. 4 are normally adopted with
either M20 or M24 bolts. It is common in the design to assume that the bending moment is
resisted by the flange plates, the shear by the web with any coexistent axial load being
divided equally between the flanges; this is illustrated in Figures 5.32 (a), (b) and (c).
Compression
·,
Mom~
Compressive
"Ol force
( oi
.~J Tensile
force
Tension
Figure 5.32(a)
Shear force
  J
Figure 5.32(b)
Fx
r , r
10 01 10 2
I 0 0 I I 0
Fx I ""
L':..
..,. I
_:_J Fx
I ..,.
L::. Fx
2
Figure 5.32(c)
Since the cover plates are the same as the beam flanges, the possibility of failure due to
strut action need not be checked; however the calculations have been carried out to illustrate
the method.
The equation given in Appendix C: Clause C. 3 : can be used to evaluate the secondary
moment
M rlfcs
max(
1
fc)
Pc
where:
Mmax is the moment about the yy axis to be shared by both cover plates
17 = 0.001a(A A.,) ~ 0
a = 3.5
A, slenderness Ljryy
s is the plastic modulus
I I
A., (~ 2 £y
0.2 0.2 (~ 2 x205x10 3 y 17.15
Py 275
17 0.001 X 3.5 X (122 17.15) 0.367
263.2 X 103
!c = 30.75 N/mm2
85.6 X 102
Syy = 187 x 103 mm3
~ 2 X 205 X 103
Pe = = 135.94 N/mm2
122 2
0.367 X 30.75 X 187 X 103
Mmax = = 2.73 kNm
( 1  30.75) X 106
135.94
= 1.37 kNm per flange
The flange cover plates should be checked for the combined axial load due to the
primary moment and applied axial load, and the secondary moment equal to Mmax I 2 due to
strut action.
F mMY
Clause 4. 8. 3. 3.1 +
AePy Znet.Py
Connections 175
where:
F the applied axial load,
m the equivalent uniform moment factor (assumed to equal 1.0 in this case),
M the applied moment about the yy axis due to strut action,
Ae the net crosssectional area of the cover plate,
Znet the elastic section modulus of the cover plate allowing for the holes,
py the yield stress from Table 6.
22 mm 22 mm
!Y
30mm 90mm 30 mm
150 mm
Figure 5.33
3.56x10 6
Znet 38.13 x 103 mm3
87.5
Anet = (150 x 15) (2x 22x 15) 1590 mm2
278.4
Minimum number of bolts required ~    3.9
71.3
Adopt eight M20 H.S.F.G. bolts with two pairs on each side
In H.S.F.G. bolted connections, the outer splice (i.e. cover plates) should not be thinner
than the lesser of:
rrtrr
I I I I
I I I I I 30
 1
90 150
cbcb +~~
I I : I I
 1
I I I I I 30
I I I I
Flange cover plate
150 mm x 320 mm x 15 mm thick
Figure 5.35
Clause 6.4.2.3: When the distance, Lj, between the first and last rows of fasteners (200 mm
in the case above), is greater than 500 mm, then the slip resistance of the bolts should be
determined using:
Connections 177
Try a plate 240 nun wide x 300 nun deep x 10 nun thick.
Clause 4.2.3
where:
Av = 0.9[(10 X300} (3 X 22}] = 2640 nun2
Pv = (0.6 X 275 X 2640)/10 3 436 kN > 125 kN
Cover plates are adequate in shear
A conservative estimate of the bending strength of the web cover plate can be made
assuming that:
where .Au used to determine Pb is calculated using the equation given in Appendix B: Clause
B.2.7,
I
Table 19
1 60mm
90
90
r=:fl
'
·r·125 kN
__17.5 kNm
f125 kN
'
41.67 kN
<p 41.67 kN G
<p 41.67 kN + 0
<p 0
41.67 kN 41.67 kN
Direct Rotational
shear shear
Figure 5.36
As with beam splices, continuity of a structure must be maintained at the location of the
splice. It is necessary to ensure adequate stiffuess about both axes in addition to ensuring
adequate strength. In braced multistorey frames, the nominal moments induced by
incoming beams (see Section 4.2 Chapter 4) as indicated in Clause 4. 7. 6 must be included
in addition to the axial load when designing the splice. The bending moments are assumed to
be carried solely by the flanges whilst the axial load may be shared between the web and the
flanges in proportion to their crosssectional areas.
Although not essential, it is usual to assume that the column ends are prepared for
contact in bearing and that untorqued bolts in clearance holes are used.
Generally column splices are situated a short distance, e.g. 500 mm above floor level. In
circumstances in which a splice is located away from a point of lateral restraint, the effects
of strut action as illustrated in Section 5. 6.1 for beam splices must be considered.
It is advisable wherever possible to minimize secondary effects due to eccentricity and
splices should therefore connect members in line such that the centroidal axes of the splice
plates coincide with those of the column section above and below the splice. Columns of the
same serial size but of different mass per metre run and of different serial sizes can be
Connections 179
accommodated by the use of packing plates, web plates and angle sections as shown in
Figure 5.37
In circumstances in which it is necessary for the splice to comply with structural
integrity requirements as specified in Clause 2. 4. 5. 3, the flange cover plates and bolts
should be capable of transmitting a tensile force equal to two thirds of the factored vertical
load applied to the column floor immediately below the splice. If there is no net tension due
to the combined moment and axial load, then the splice can be designed to transmit the
applied compressive load by bearing.
The empirical detailing requirements indicated should ensure that sufficient stiffuess and
robustness of a splice is achieved when only bearing is being considered. If net tension does
exist when considering combined moment and axial loading then the strength of the flange
cover plates must be checked in tension and bearing, and the bolt group checked in shear.
Recommended splice details to ensure adequate stiffuess and robustness and continuity.
+ The thickness of the flange cover plates should be at least equal to the greater of
10 nun or half the upper column flange thickness.
+ When cover plates are positioned on the outside, there width should be at least
equal to the width of the upper column flanges.
+ The length of the cover plates should be at least equal to the greater of
(i) 225 mm + division plate thickness or
(ii) the width of the upper column flanges
+ When colunms of the same serial size are used a nominal web cover plate of width
at least equal to half the overall depth of the upper column should be used.
+ When colunms of different serial sizes are used, a division plate of sufficient
thickness to allow a 45° dispersion of load from the upper column to the lower
column should be used in conjunction with a pair of nominal web cleats.
180 Design of Structural Steelwork
+ Grade 8.8, untorqued bolts are normally adopted with typically 4 No. M20 bolts in
the web plates.
+ When combined bending and compressive loading results in a net tensile force being
induced, then torqued H.S.F.G. bolts should be used with the cover plates being
checked for bearing and tension.
+ The spacing of the flange cover plate bolts perpendicular to the direction of stress
should be maximized to optimize the joint rigidity.
A column in a multistorey frame is spliced above the first floor level as shown in Figure
5.38. Design a suitable splice which must satisfy structural integrity requirements in
addition to transferring the given ultimate loading.
I
I"'
t Fz 1
t
3.5 m
f 203 X 203 X 52 U C
:z: I
t F1
3.5 m
I Splice position
I :z: 1
r 254
t F1
X 254 X 73 UC
t
_l
3.5 m
Figure 5.38
5259 11410
Clause 4. 7. 4
(~)upper 
350
= 15
(~)lower =
400
= 28.5
(~)lower 28.5
= 1.9 ~ 1.5
(~)upper
15
Clause 4. 5.1. 3 Assume a 45° dispersion of the load through the division plate
Thickness of plate required ~ D1  Dz 254  206 = 24 mm
2 2
Adopt division plate 25 mm thick
Check for the presence of tension in the splice due to combined axial load and bending
moment
Axial load induced in cover plates by the moment ~ 342 X 103 = 1659 kN
206.2
Since this value is greater than the applied compressive force of 1507 kN tension is
developed and it is necessary to check the tensile and bearing capacities of the flange cover
plates and the shear capacity of the bolt group. In addition, Clause 6.3.4 relating to long
joints must be satisfied.
s gross area
grossarea = 205 x 10 = 2050 mm2
netarea 2050  (2 x 22 x 10) = 1610 mm2
Ke x netarea 1.2 x 1610 = 1932 mm2
Pt = (1932 X 275)/10 3 = 531 kN
20 kN
(iv)
Clause 4. 6.1 Plate tension capacity P 1 = Aepy
Table 33 Ae = KeAne! < Agross
Table 6 Ke = 1.2, py = 275 N/mm2
Anet = (50 18) X 10 = 320 mm2
Agross = 50 X 10 = 500 mm2
Ae = 1.2 X 320 = 384 mm2
45 mm 50 mm 45
rtm I I
Contract: Connections Job Ref. No.: Example 5.2 Cales. by: W.McK.
Part of Structure : Simple Bolted Connections Checked by:
Calc. Sheet No. : 2 of 2 Date:
Clause 4.6.1
Clause 3.3.3
Table 6 Anet = (130 44)12 = 1032 nun2
Agrass = 130 X 12 = 1560 nun2
A. = (1.2 x 1032)/103 = 1238.4 nun2
Plate tension capacity = (1238.4 x 275)/10 3 = 340.6 kN
(J
Factored design load 50 kN/m ,
f=l
Bolt Group A: M20 Grade 8.8 bolts in double shear
50 (j.i
Contract: Connections Job Ref. No. :Example 5.5 Cales. by: W.McK.
Part of Structure : Simple Bolted Connections Checked by:
Calc. Sheet No. : 2 of 9 Date:
9
9
9
9
Vertical shear force /bolt = 175 kN = 43.75 kN
4
Rotational shear force I bolt =
~37.5~
 r
,...
40
~40




=l 290 mm
v:
~t/, 
=J
  
  ~
'
t%
Cleat Bearing
Table 33 Bearing Capacity Pbs = dtpbs _::: 0.5etpbs
e = 40 since e = 2d dtpbs = 0.5etpbs
Pbs = 460 Nlmm2 Cleat is adequate
Pbs= (20 X 10 X 460)/10 3 = 92 kN > 57.612 = 28.8 kN bearing
T~~
42mm ~~
l~~
i web thickness = 11.4 mm
r
l10mm
190 Design ofStructural Steelwork
! !
4 4r
yr
1 1
qr
4 4r
v
I I
yr
i i
Bolt Bearing
Clause 6.3.3.2 Pbb = dtPhb where [Jbb = 460 Nlmm2
Table 33 Pbb = (20 X 10 X 460)/10 3 = 92 kN Bolts are adequate
pbb > 21.9 kN in bearing
Cleat Shear
Shear Force/ Cleat = 175 I 2 = 87.5 kN
Clause 4.2.3 Pv = 0.6pyAv
Table 6 py = 275 Nlmm2 ; Av = 1818 mm2 Cleats are adequate
Pv = (0.6 X 275 X 1818)/10 3 = 300kN in shear
Cleat Bearing
Clause 6.3.3.3 Pbs = dt[Jbs ~ 0.5etpbs
Connections 191
Contract : Connections Job Ref. No. : Example 5.5 Cales. by: W.McK.
Part of Structure : Simple Bolted Connections Checked by:
Calc. Sheet No. : 5 of 9 Date:
Table 33 As before e = 2d
Pbs = (20 X 10 X 460)110 3 = 92 kN Cleats are adequate
Pbs > 21.9 kN in bearing
I I
i i
Bolt Shear
Clause 6.3.2 Shear capacity in single shear Ps = 91.9 kN
Table 32 Maximum design shear force/bolt = 17 5 I 8 = 21.7 5 kN Bolts are
Ps > 21.75 kN adequate in shear
Bolt Bearing
Clause 6.6.3.2 Bolt bearing capacity Pbb = (20 x 8 x 460) I 103 Bolts are
= 73.6 kN > 21.75 kN adequate in bearing
Plate bearing
Clause 6.3.3.3 Pbs = dtpbs :::; 0.5efpbs e = 2d
Table 33 Pbs = (20 X 8 X 460) I 10 3 = 73.6 kN > 21.75 kN Plate adequate in
bearing
Web shear
175 kN
~~
290mm
0
__l
p0'
~40~
Pv = 0.6pyAv

Av = (0.9 X 290 X 11.4) = 2975 mm2
Pv = (0.6 X 275 X 2975) I 103 = 491 kN > 175 kN Beam web is
adequate in shear
Clause 6.6.2 Weld
Assuming 6 mm fillet welds, E43 electrodes
Effective length of weld = 2 [290 ( 2 x 6)] = 556mm
Clause 6.6.5 Pw = 215 Nlmm2 throat size = 0.7 x 6 =4.2mm 556 mm length
Table 36 Strength of weld I mm = (4.2 x 215) I 103 = 0.903 kN 6 mm fillet weld is
Weld capacity = 0.903 x 556 = 502 kN > 175 kN adequate
40
[]~~
 1 10mm
60 ~40
100
~
~10mm
*
11.4mm
Connections 193
Contract :Connections Job Ref. No. : Example 5.5 Cales. by: W.McK.
Part of Structure : Simple Bolted Connections Checked by:
Calc. Sheet No. : 7 of 9 Date:
_d
Pbs _ 20 X 10 X 460 == 92 kN
Table 33 lPbs 3
10
10.5 X 10 3 X 105
Rotational shear force I bolt ==
2( 35 2 + 105 2 )
== 45 kN
== J43.75 2 +452
= 62.8 kN Web and Plate are
Pbs> 62.8 kN adequate in bearing
e 1 5. 2.5d == 2.5 x 20 == 50 mm
::; e1 = 50 + 40 == 90 mm :. e1 ==50 mm
194 Design of Structural Steelwork
Contract : Connections Job Ref. No. : Example 5.5 Cales. by: W.McK.
Part of Structure : Simple Bolted Connections Checked by:
Calc. Sheet No. : 8 of 9 Date:
Clause B.2. 7
I
LEd)z
A.LT =nx2.8x (7
LE=60nun
1175 kN
~~
290 mm
0
_j_
1'"o__J
~40~
L
I
60x290)2
ALT = 0.77 X 2.8 X ( 102 = 28.4
Connections 195
Contract: Connections Job Ref. No.: Example 5.5 Cales. by: W.McK.
Part of Structure : Simple Bolted Connections Checked by:
Calc. Sheet No.: 9 of 9 Date:
(10.5 X 10 3) X 145
Max. bending force/rum= 6 = 0.38 kN/m
4.06 X 10
Contract : Connections Job Ref. No. : Example 5.6 Cales. by: W.McK.
Part of Structure : Bolted Moment Connections Checked by:
Calc. Sheet No. : 1 of 2 Date:

·' ~:">.'''

Ultimate design moment = 540 kNm
 ~
. Moment
Flange force mduced by the moment = Ft =
lever arm
F, _ 540 X 10 6 = 915 .4 kN
t (609.619.7)
Moment:
Tension capacity P1 = 0.9Po ;::: 915 ·4 = 229 kN
4
. . .
Clause 6442 Min. . d
P0 requrre 229
= ·0 = 2554kN
. Adopt 4 No. 30 mm
0.9
For 30 mm H.S.F.G Bolts Po = 286 kN H.S.F.G. Bolts
Shear:
Slip resistance PsL = 4 x (1.1 x KsJ.IPo) ;::: 380 kN
where Ks = 1.0 and JJ. = 0.45 Bolts are adequate
PsL provided = 4 X 1.1 X 0.45 X 286 = 566 kN ;:=: 380 kN in shear
Contract: Connections Job Ref. No.: Example 5.6 Cales. by: W.McK.
Part of Structure : Simple Bolted Connections Checked by:
Calc. Sheet No. : 2 of 2 Date:
Clause 6.4.2.2 Clearly in this case the shear and bearing capacity of the
connection are more than adequate.
6.1 Introduction
In many instances, it is necessary to support heavy vertical loads over long spans resulting
in relatively large bending moments and shear forces. Ifthe magnitude ofthese is such that
the largest Universal Beam section, even when compounded with plates, is inadequate then
it is necessary to fabricate a beam utilizing plates welded together into an I shaped section.
The primary purpose of the flange plates is to resist the tensile and compressive forces
induced by the bending moment. The primary purpose of the web plate is to resist the
shearing forces. These sections are normally more efficient in terms of steel weight than
rolled sections, particularly when variable depth girders are used, since they can be custom
designed to suit requirements. In addition, the development of automated workshops has
reduced fabrication costs, particularly when compared to box girders and trusses which are
still fabricated manually resulting in high costs.
When compared to other forms of construction, plate girders do have some dis
advantages, e.g. they are heavier than comparable trusses, more difficult to transport, have
low torsional stiffuess, can be susceptible to instability of the compression flange during
erection and do not easily accommodate openings for services.
Since plate girders are normally highly stressed, the possibility of local buckling will tend
to be higher than in rolled sections, particularly with respect to the web plate. In most cases,
stiffeners are provided along the length of the girder to prevent the web plate from buckling.
A typical plategirder and its component parts is shown in Figures 6.1 and 6.2; in
addition a flow chart is shown in Figure 6.3, indicating the principal steps which are
undertaken when designing such a girder.
'TFillet
~Welds
Figure 6.1
198
Plate Girders 199
l modifytria1 section
!Initial sizing 1
suitable l L unsuitable
!Determine moment capacity I
suitable l I
unsuitable
!Determine shear capacity I
suitable l I
unsuitable
!check deflection
suitable
1
!Design intermediate transverse stiffeners I
suitable
1
!Design load bearing stiffeners I
suitable l
!Prepare detail drawing of girder I
Figure 6.3
200 Design ofStructural Steelwork
6.1.1 Design Load Effects
The design load effects on a plategirder are determined using standard structural analysis
techniques. They comprise the effects induced by bending which are primarily,
Bending
Moment
( ~/
F;/ Figure 6.4
(i) Compressive and tensile forces above and below the neutral axis of the crosssection
Figure 6.5
(ii) Shearing forces which are resisted by the web
Figure 6.6
(iii) Forces tending to induce failure by web buckling
Figure 6.7
(iv) Forces tending to induce failure by web crushing
(v) Vertical deflection
The magnitudes and senses of these effects are dependent on the magnitude and nature of
the forces applied to the girder. The service design loading for buildings is given in BS
6399:Part 1 and the appropriate load factors can be found in Table 2 ofBS 5950:Part 1.
Plate Girders 201
The initial sizing of the component parts of a plate girder is carried out on the basis of
experience and consideration of the section classification as given in BS 5950:Part 1. While
webs tend to be slender, they should not be excessively so, since additional stiffening will be
required with its inherent fabrication costs.
Tables 6.1 and 6.2 given below indicates typical values of spantodepth ratios and web
thicknesses for plate girders used in buildings.
Typical Typical
spanto thickness
Applications depth Beam depth (mm) of web
ratio (mm)
1 simplysupported, non I
(i) : composite girders with I
I
I concrete decking; I
I I
1 constant depth beams used 1220 (a) 1 Up to 1200 10
I
: in simplysupported I
1 composite girders I
I
: constant depth beams used I
(ii) : in continuous non I
I
1 composite girders using 1520 (b) I 1200 1800 12
I
: concrete decking I
1 simplysupported crane I
I
(iii) : girders 105 (c) I 1800 2250 15
I I
I
I
(d) : 2250  3000 20
I
The moment capacity of plate girders where the flanges of the section are plastic, compact
or semicompact, but where the web is slender (as defined in Table 7) or is thin and
susceptible to shear buckling (i.e. d/t ~ 63s), can be evaluated using one of three methods
outlined in Clause 4.4.4.2 ofBS 5950:Part 1. Method (a) in which the moment is assumed
to be resisted by the flange plates alone subjected to a uniform stress of py and the shear
stress is assumed to be resisted by the web is adopted in this text. This method is
considerably easier to understand and use than the other two that are given.
202 Design ofStructural Steelwork
The implication in the code of using a uniform stress equal to py, is that the compression
flange is fully restrained, In instances where this does not occur, a value of Pb can be
determined using the equivalent slenderness ~T and Table 12. The equivalent slenderness
can be evaluated using Clause 4.3. 7 or the equations given in Appendix B of the code.
For the design of sections in which the flanges are slender, a reduction factor should be
applied to the uniform stress value; this can be found in Table 8 of the code.
The section classification of the web as indicated in section 2.2 relates to local buckling
induced by longitudinal compressive stresses. Webs can also fail due to local buckling
caused by concentrated loads applied to the flanges, as indicated in Section 2.5.1. In
addition to these two modes of failure, a thin web can fail owing to shear buckling in which
diagonal compressive and tensile stresses are developed within web panels between
stiffeners. The compressive stresses induce outofplane shear buckling and tensionfield
action subsequently enhances the loadcarrying capacity until failure occurs with the
development of plastic hinges in the flanges, as shown in Figures 6.8 and 6.9.
shear
force
shear shear
force force
shear
force
Figure 6.8
L Plastic hinge
Figure 6.9
The code permits allowance for tensionfield action in the design of webs and stiffeners.
This technique enables a more accurate assessment of the contribution made by stiffeners to
the strength of the web and hence a more economic section to be designed. Tensionfield
action is not considered in this text.
The shear capacity of a section when tensionfield action is not being considered is given
in Clause 4.4.5.3. This is dependent on the critical shear stress qcr, given in Table 21 and
will determine if intermediate stiffeners are required.
Plate Girders 203
6.4 Deflection
As with Universal Beam design, the deflection induced by the unfactored imposed loads
should not exceed the limits specified in Table 5 of BS 5950:Part 1. The actual deflection
can be calculated using coefficients or the equivalent uniform load distribution technique,
both of which are illustrated in the example at the end of this chapter.
Since, for economy, plate girders are designed with thin webs (i.e. Clause 3. 6. 2 BS
5950:Part 1, dlt 2:. 63&), shear buckling must be considered. The serviceability
requirements which require a minimum thickness to prevent buckling of the web and of the
flange into the web are given in Clauses 4.4.2.2 and 4.4.2.3 ofBS 5950:Part 1 respectively.
As mentioned previously, the shear buckling resistance of webs can be calculated from
Clause 4. 4. 5. 3. If this value is exceeded, then it will be necessary to include intennediate
stiffeners to prevent outofplane buckling of the web.
When a concentrated load is applied through the flange, for example at a point of support or
under a column, it is sometimes necessary to provide loadbearing stiffeners, which may be
single or double, to prevent local buckling and crushing of the web. The load carrying
capacity of the web in buckling and bearing can be determined using Clauses 4. 5. 2 and
4.5.3 (see Chapter 2 Section 2.5 for universal beam design).
Where loadbearing stiffeners are provided at supports they are also known as end posts.
The design of end posts including their connection to the web and flange plates is set out in
Clauses 4. 5. 4 to 4. 5.11. In some instances it is necessary to check the compression edge of
the web between stiffeners; the procedure for this is given in Clause 4.5.2.2 and is
illustrated in Example 6.2 of this chapter.
I' I • Roof
I I
I I I
I I I I I
I
I I
t
'I
I
I
I
I
I
I
I
l
I
I
I
I
I
3rd. Floo
'I I
I
~
I
f
I
I
I
2nd. Flo or
I I I I
I
t
I
1st. Floo
±11 I I I I I I I I
I I
I L PIa te girder
I
I I
~ ~?0 ~ ~~
3.0 m 3.0 m I 3.0 m I 3.0
.....................................................................................................................................
Design Data:
Characteristic dead load on the roof (including. selfweight) = 4.0 kN/m2
Characteristic dead load on the floors (including. selfweight) = 5.0 kN/m2
Estimated selfweight of the plate girder 4.5 kN/m
Characteristic imposed loads on the roof and floors as per BS 6399:Part 1
Assume that the compression flange is fully restrained throughout and that Grade 43 steel is
used.
Plate Girders 205
6. 7.1 Design Loading
The design loading on the girder is determined using static analysis in which the floor and
roof beams are assumed to be simply supported. This results in the girder being subjected to
a uniformly distributed load due to the selfweight, and three concentrated loads at the
internal colunm positions.
The code of practice for the dead and imposed loading on buildings (BS 6399:Part 1)
permits a reduction in the value of the imposed loads on multistorey structures. In cases
where a single span of a beam or girder supports not less than 40m2 of floor area at one
general level, the reduction is in accordance with Table 3.
In this design the area of floor supported is equal to (7.5 m x 12.0 m) = 90 m2 and
therefore the imposed load should be reduced accordingly, i.e. using linear interpolation and
the values given in Table 3 the reduction equals 6.25%. It should be noted that the Table 3
values do not apply to roofloadings (Clause 5.2).
A typical internal colunm supports the roof/floor area shown m Figure 6.11, 1.e.
(7.5 x 3.0) = 22.5 m2
,,, '''l
m x 3.0 m =~=i:l
I
Figure 6.11
6.3 kN/m
1396.2 kN 1396.2 kN
Bending
Moment
Figure 6.12
an estimate of the flange force and hence the flange area required can be made.
Plate Girders 207
::·:·:·:·:·:·:·:·:·:·: :.:·:.: 50 mm
10 mm 1000 mm
Figure 6.13
The crosssectional area ofthe proposed girder 2 x (450 x 50)+ (1000 x 10)
55 x 103 mm2
Unit weight of steel = 7. 85 x 108 kN/mm3
This is slightly less than the selfweight assumed in the initial sizing of the girder.
The local buckling of crosssections can be avoided by limiting the width to thickness ratios
of individual elements which are subjected to compression. The various elements of a cross
section, predominantly the flanges and webs, are classified in BS 5950:Part 1, Table 7 as
being plastic, compact, semicompact or slender, the definitions of which are given in
Clause 3.5.2. Where elements of a crosssection are shown to be slender, then additional
measures which must be taken to prevent local buckling are given in Clause 3. 6.
6. 7. 5 Flanges
6. 7.6 Web
Clause 4. 4. 2. 2
To ensure that the web will not buckle under normal service conditions it is necessary to
satisfy the following minimum web thickness criteria:
d
(a) without intermediate stiffeners: t ~
250
(b) with transverse stiffeners only :
d
(i) where stiffener spacing a > d then t ~
250
(a)2
1
2
(ii) where stiffener spacing a ::;; 1.5d then t 2: d ( Pyr)
250 455
Assuming the more critical case with stiffeners, i.e. a 2: 1.5d,
then t 2: 1000(255) = 3.0 mm
250 345
Clearly the web is satisfactory in both respects.
The moment carrying capacity of the section ignoring the web is given by :
M= pyf X Sx:x
where:
Sx:x is the plastic section modulus based on the flanges alone and equals (flange area x
distance between the centroids of the flanges)
Sx:x = (450 x 50) x (1050) = 23.625 x 106 mm3
M = 255 x 23.625 = 6024 kNm.
M applied < M therefore the section is adequate with respect to bending
Note: If the compression flange had not been fully restrained, then the value of pyf would
have been replaced by a value of Ph from Table 12. This is considered at the end of this
chapter in Example 6.2.
Designing without tensionfield action, the shear capacity of the section is given by:
where qcr is the critical shear strength obtained from Tables 21 (a) to (d) as appropriate. If
no intermediate stiffeners are assumed the spacing is taken as infinity.
The above equation can be rearranged such that
       \
      
115 165 165 165 154 142 132
120 165 165 162 149 137 126
125 165 165 158 144 131 120 )
__.)
.....___..    
,l 4@ 750 crs. ~" 4@ 750 crs. ,l 4@ 750 crs. ~, 4@ 750 crs. ,,1._
Figure 6.15
In the case of a girder supporting a larger uniformly distributed load the shear force may
reduce significantly towards the centre and hence permit an increase in the stiffener spacing
towards the middle of the span.
6. 7. 9 Deflection (Clause 2. 5)
Clause 2. 5.1
In Table 5 the deflection limit for beams without plaster or brittle finishes is given as
0 < span
max  200
This is the deflection induced by the unfactored imposed load only, i.e. the service
imposed load.
Plate Girders 211
The actual deflection can be determined using appropriate coefficients such as those
given in Table 2.1 in Chapter 2, or can be estimated using the equivalent uniform load
technique.
Clause 3.1.2
Table 6 the modulus of elasticity E == 205 kN/mm2
Service load I column == (33.75 + (3 x 52.73)) == 191.9 kN
The midspan deflection for a beam with three point loads can be evaluated by super
imposing the contributions from each load i.e.
,
(j ~2
{191.9 (12000) 3 [ 3 x 3  4
X
3
( 3 ) ]}
 + 191.9 X (12000)
3
Jmax ~ 6.43 mm
Note: The 2 above can be used in this instance because of the symmetry.
12000
Table 5 value =:: == 60 mm
200
287.9kN t t 287.9kN
3.0 m I• 3.0 m • I• 3.0 m , I , 3.0 m ·I
12.0 m .1
Figure 6.16
0.104 B. Max L2
The estimated maximum deflection is given by
EI
In this case
Clause 4. 4. 6. 3
The maximum outstand (b,) of intermediate stiffeners is governed by the requirements of
Clause 4. 5.1. 2, i.e.
When the outer edge is not continuously stiffened then b, :s; 19t,e
and when 13t,e :s; b, :S;19t,e the design should be based on a core with b, = 13t,e
Assuming 8 mm flats for the stiffeners and e = 1. 0 then b, :s; 19 x 8 = 15 2 mm
Clause 4. 4. 6. 4
The minimum second moment of area, /,, about the centreline of the web should satisfy:
I, ? 0.75cf for a ? J2d
and
1.5d 3 t 3
I, ? for a < J2d
a2
t 8 (2bs +
12
tf
flange
plate
t b, )
t II 2b, + t
b,
(2bs + t) == 158 mm
~ \
(15810) Section through stiffeners
outstand = 74mm and web
2
say 75 mm Figure 6.17
Clause 4. 4. 6. 7
The weld between the web and intermediate stiffeners not subject to external loads should
10 2
:. weld capacity ~ = 0.27 kN/mm run on two welds
5 X 75
It is normal practice to use a minimum fillet weld in structural elements of this type of not
less than 6 mm, (strength= 0 .9 kN/mm).
Adopt four continuous 6 mm fillet welds for each set of intermediate stiffeners. Since the
stiffeners are not subject to external loading, they may terminate clear of the tension flange
by an amount approximately equal to 4t, i.e. 40 mm.
The stiffeners should extend to the compression flange but no welding is required at this
location.
Clause 4. 5. 4. 2
The minimum area of stiffener which should be in contact with the flange to limit bearing
stresses is determined by :
(0.8 X 1396.2 X 10 3 )
A ~ = 4061 mm2
275
Figure 6.18
Clause 4.5.1.2
The maximum outstand requirements are the same as those for the intermediate stiffeners.
13t,B = 13 x 12 x 1. 0 = 156 mm
19t,c = 19 x 12 x 1. 0 =228mm
214 Design ofStructural Steelwork
since 13ts& :::; bs :::; 19t.s the stiffeners should be designed on the basis of a core section
equal to 1318 &, i.e. 156 mm.
Clause 4.5.1.5
The buckling resistance of the stiffener is considered to be the compressive resistance of a
stru~ with a crosssection equal to the full or core area of the stiffener together with an
effective length of web on each side of the centreline of the stiffeners, (where available},
limited to 20t. This is shown in Figure 6.19.
The compressive strength Pcis taken from Table 27(c) which requires a slenderness ratio
)., = effective length
radius of gyration
The radius of gyration of the crosssection should be taken about an axis parallel to the web
ryy = g =
33.4 X 106
5804
= 78.56mm
T·.; 12 mmthick
B
:\0
Figure 6.19
The compressive strength Pcis taken from Table 27(c) which requires a slenderness ratio
)., = effective length
radius of gyration
The radius of gyration of the crosssection should be taken about an axis parallel to the
web
12 X 3223
I = +
12
Area = (322 x 12) + (194 x 10} = 5804 mm2
Plate Girders 215
~=
33.4 X 106
ryy = = 75.86mm
5804
The effective length for load bearing stiffeners where the flange through which the load is
being tra.ilsferred can be considered restrained against lateral movement relative to the other
flange should be considered as:
(a) Le = 0.7L
where restraint to the flange against rotation in the plane of the stiffener is provided by other
structural members,
(b) Le =L
where no such restraint is provided.
(The effective length for intermediate transverse stiffeners should be assumed to be 0. 7L.)
In each of the cases above Lis the length of the stiffener.
Assuming case (a) above;
700
Le = 0.7 X 1000 700 mm; A = 9.23
75.86
Clause 4. 7. 5
The design strength obtained from Table 6 should be reduced by 20 N/mm2 for welded
sections, therefore since t < 16 py = 255 N/mm2 •
From Table 27(c) when A= 9.23 and py= 255 N/mm2 Pc = 255 N/mm2
The buckling resistance ofthe stiffener Px = (255 x 5804)/10 3 = 1480 kN
Since Px ;;:: 1396.2 kN the stiffeners are adequate with respect to buckling
Clause 4. 5. 5
The bearing capacity of the stiffeners should be sufficient to sustain the difference between
the applied load or reaction and the local capacity of the web, as given in Clause 4. 5. 3. i.e.
where:
b1 is the stiff bearing length as defined in Clause 4.5.1.3 and Figure 8 of the code
n2 is the length obtained by dispersion of the load through the flange to web connection
at a slope of 1:2.5 to the plane of the flange.
t is the web thickness
pyw is the design strength of the web from Table 6
J Web !
12 mm /.t
j
l I Bearing length
1\ I
Bearing area
50 m m t=" ·: .,:~:;:;;::~~=:~:~:;:,:,~:::i::~::::~:::::~:::~:~:::~:::::t:':':~:~:#(:~:::::::~(:~::f:::::t:: I
u
.+ 2.5 x 50= 125 mm , n2 = 131 mm Fla~ge plate
1396.2 kN \I
I '\
:" Longitudinal section through web
Figure 6.20
since this is greater than 1036 kN the stiffeners are adequate in bearing.
Clause 4. 5. 9
The connection of the bearing stiffeners to the web should be designed to resist the lesser of:
(5804 X 275)
Pt = kn..
ory
= I 03 = 1596 kN
or
(ii) the larger of the sum of any forces acting in opposite directions or the sum of any
forces acting in the same direction, i.e.
Figure 6.21
The connection between the flange and the web must be designed to transfer this
horizontal shear stress 'q'. The value of' q' can be determined from
QAY N/mm2
q = __
It
horizontal
shear stress
Section
considered
Q!AY
= __
Shear force I nun length = qx t Nlnun
I
where:
Q is the vertical shear force at the position along the span being considered,
Ay is the first moment of area (about the neutral axis) of the part of the crosssection
above the level at which the shear stress is being evaluated,
I is the second moment of area of the crosssection about the neutral axis.
Note: An intermittent weld could have been used here but this tends to create potential
corrosion pockets and is best avoided.
A welded plate girder is simplysupported over a span of 15.0 m and supports two columns
at the thirdspan points in addition to a uniformly distributed load along its length. Adequate
lateral restraint is provided to the compression flange at the supports and column locations.
Using the data provided design a suitable plate girder.
Design data:
Column loads
service dead load = 140kN
service imposed load = 400kN
Uniformly distributed load
service dead load = 8.0kNim
service imposed load = 25.0kNim
Assumed selfweight of girder = 3.0 kNim
Contract: Connections Job Ref. No.: Example 6.2 Cales. by: W.McK.
Part of Structure : Plate Girders Checked by:
Calc. Sheet No. : 1 of 11 Date:
836 kN 836 kN
1251.5kN
~
~t1251.5kN
1: ·I 1s.om I :1
For analysis this can be considered as the superposition
of two separate load cases as shown below :
~4~/m
415.5 kN t~ 415.5 kN
1: • 15.0 m • :I
836 kN 836 kN
1251 .5kN
Shear Force Diagram
5738 X 10 3
Flange force = 4781 kN
1200
Clause 4.3. 7.3 Since the flange is not fully restrained a value less than
265 N/mm2 should be used when estimating the required
flange area. The moment capacity will be based on:
10mm
1200mm
1 460mm j
ryy =
Fi= 649
48.8
X
X
10 6
10 3
= 115.32 mm
Contract: Connections Job Ref. No.: Example 6.2 Cales. by: W.McK.
Part of Structure : Plate Girders Checked by:
Calc. Sheet No. : 4 of 11 Date:
Web:
1
(a)2
1
A = 48.8 x 10 3 mm2
hs = 1200 + 40 = 1240 nun
u =
4 X (22.82 X 10 6 r X 0.985]0.Z
5

[
(48.8 X 10 3
I
r X 1240 2
0.859 u = 0.859
X = 0.566 X 1240 X (
48·8 X
103 ) 2 = 34.64 X= 34.64
20.03 X 10 6
Clause 4.3. 7.6 Assuming that the plate girder is not subject to
destabilising loads
Table 13 m = 1.0
Table 17 {J is +ve. and y lS +ve
55.4 X 52
Mo= = 173.125
8
M 5565
Table 16 r = 173.125
= 32.14 n = 0.98
Contract: Connections Job Ref. No.: Example 6.2 Cales. by: W.McK.
Part of Structure : Plate Girders Checked by:
Calc. Sheet No. : 6 of 11 Date:
Table 16 For py = 265 Nlmm2 and ..1u = 35.77 Ph= 263 Nlmm2
Mb=Sxx·Pb
Mb = (263 x 22.82 x 10 6 )110 6 = 6001.7 kNm. Section is adequate
Mb>Mx with respect to
bending
Clause 4.4.5.3 Shear Capacity :
Outer third of span
Vcr = qcr<it where Vcr = 1251.5 kN
15000 3 [5 X 25 X 15 + 23 X 400]
205 X 15.59 X 10 9 384 648
Contract : Connections Job Ref. No. : Example 6.2 Cales. by: W.McK.
Part of Structure : Plate Girders Checked by:
Calc. Sheet No. : 8 of 11 Date:
1.5 X 1200 3 X 10 3 6 4
= 1.659 x 10 mm
1250 2
I. = 1.659 X 106
0.8 X 836 X 10 3
Clause 4.5.4.2 Contact area A >  = 2432 mm2
275
Clause 4.5.1.5
:!1 10 mm thick
·I
l
·w~...~
f=.==Elil=···=~·>··=~ · ··>1 :·:m:;!
·:····ill
·nill:·:·ill··=ill ·'·"ill···ill·
······ill ·"'ill···mill
:···ill ·< C::=~ 10 mm thick
· i:i!iJ
!! (
!
'     1              1
201 = 200 mm i 201 =200 mm
I I
J0 X 270 3 400 X 10 3
I = + = 16.44 x 106 mm4
12 12
Area = (270 X 10) + (400 X 10) = 6700 mm2
/16.44 X }0 6
r = = 49.5 mm
6700
Clause 4.5.1.5
\
J
rlI
322mm
1
••
.x«··
~~
12 mm thick
· ""' ' ··
10 mm thick
1
I 1
~~
!I 2ot =200 mm
1<:::: 'f,
10 322 3 206 X 10 3
X
I = + = 33.41 x 106 nun4
12 12
Area = (322 X 10) + (246 X 10) = 6324nun2
/33.41 X J0 6
r = = 72.7nun
6324
Plate Girders 229
1251.5 11.408 I 06
=
X X
q = 0.915 kN/nun 6 nun fillet welds
15.59 X 10 9
7. Concise Eurocode 3 CEC3
7.1 Introduction
This chapter provides an introduction to the contents of the Concise Eurocode 3 and
illustrates the design of simple structural elements. A more comprehensive treatment can be
found in Introduction to Concise Eurocode 3 (CEC3)  with Worked Examples published
by The Steel Construction Institute.
The European Standards Organisation, CEN, is the umbrella organisation under which a
set of common structural design standards (e.g. ECl, EC2, EC3, etc.) have been developed.
The Structural Eurocodes are the result of attempts to eliminate barriers to trade throughout
the European Union. Separate codes exist for each structural material, including EC 3 for
steel. The basis of design and loading considerations is included in EC 1.
Each country publishes its own European Standards (EN); for example, in the UK the
British Standards Institution (BSI) issues documents, which are based on the Eurocodes
developed under CEN, with the designation BS EN.
Currently, the Structural Eurocodes are issued as Prestandards (ENV) which can be
used as an alternative to existing national rules. In the UK the BSI has used the designation
DD ENV; the prestandards are equivalent to the traditional 'Draft for development'
Documents.
In the UK the Eurocode for steelwork design is known as 'DD ENV 199311 Eurocode
3: Design of steel structures: Part 1.1 General rules for buildings' (together with the United
Kingdom National Application Document).
Eurocode 3 adopts the 'Limit State Design' philosophy as currently used in UK national
standards.
Each country which issues a European Standard also issues a NAD for use with the EN.
The purpose of the NAD is to provide information to designers relating to product standards
for materials, partial safety factors and any additional rules and/or supplementary
information specific to design within that country.
In the UK The Steel Construction Institute has published CEC3  Concise Eurocode 3 for
the Design of Steel Buildings in the United Kingdom. This document is an abridged version
of DD ENV 199311 and covers only the sections and Clauses which are necessary for the
design of types of steel structures most commonly used in the UK. This can be used as a
standalone design standard for most building structures. Unlike EC 3, many of the design
230
Concise Eurocode 3 CEC3 231
provisions in this document are presented in Tabular or Chart format which will be familiar
to UK designers. A summary of the abbreviations is given in Table 7 .1.
Table 7.1
Abbreviation Meanin2
CEN European Standards Organisation
EC Eurocode produced by CEN
EN European Standard based on Eurocode and
issued by member countries
ENV Prestandard of Eurocode issued by member
countries
DDENV U.K version ofPrestandard1 (BSI)
NAD National Application Document issued by
member countries (BSI)
CEC3 Concise Eurocode (abridged version of DD
ENV 199311) published by the SCI for UK
designers
The terminology, symbols and conventions used in EC 3 differ from those used by BS
5950:Part 1. The code indicates 'Principles' which are general statements and definitions
which must be satisfied and 'Rules' which are design procedures which comply with the
principles. The rules can be substituted by alternative procedures provided that they can be
shown to be in accordance with the principles; CEC3 does not make this distinction.
There are two types of Annexe in EC3: normative and informative. Normative Annexes
have the same status as the main body of the text, while Informative Annexes provide
additional information. The Annexes generally contain more detailed material or material
which is used less frequently.
Standard ISO practice has been adopted in representing a decimal point by a comma, i.e.
5,3 = 5.3.
7.2.2 Actions
7.2.3 Resistance
7.2.4 Subscripts
Multiple subscripts are used to denote variables and are defined in Appendix B of CEC3.
The level of sophistication is above that used in BS 5950:Part 1; for example, Vp1.Ra is the
design (d), plastic (pl), shear(V) resistance (R) of a section.
The term 'design' is used for factored loading and member resistance.
Design loading (Fa) = characteristic (k) value x partial safety factor (YF)
Note: Design loads apply to both serviceability and ultimate limit states.
The application of design loads to a structure results in design effects such as internal forces
and moments denoted by Sa .. Structures must be designed such that
Sa s Ra
The Eurocode provides indicative values for various safety factors and are shown in the text
as 'boxed values', i.el1,351. Each country defines 'boxed values' within the NAD document
to reflect the levels of safety required by the appropriate authority of the national
government; in case of the UK, the BSI.
Concise Eurocode 3 CEC3 233
A comparison between the values of partial safety factor for loads which are adopted in
CEC 3 and BS 5950:Part 1 is given in Table 7.2.
Table 7 2
Load Type CEC 3 BS 5950:Part 1
Permanent (G) ra = 1.35G YF= 1.4G
Variable (Q) YQ = 1.5Q YF= l.6Q
(Wind) includes wind load YF= l.4W
Permanent+ Variable 1,35G + 1,35Q 1.2G +1.2Q
(+Wind) + (0,9 X 1,35)Qwind + l.2W
The UK NAD includes a rule (Clause 4(b)) which states that the wind loading should be
taken as 90% ofthe value obtained from CP 3: Chapter V: Part 2: 1972. This compensates
for the higher value of YQ used in CEC 3. Note: CP 3: Chapter V: Part 2: 1972 has been
superseded by BS 6399: Part 2 1995.
When considering combinations of loads consisting of more than one variable load, EC3
adopts a method utilizing a combination factor If/. The values of If/ for different types of
variable load are given in ECl, but are currently included in the UK NAD. For most
building structures in normal design situations, a simplified approach is also given in EC3.
This is reflected in the modified partial load factor of 1,3 5 for the combination of more than
one independent variable load. In addition to the above, imposed floor loads and imposed
roof loads are considered to be independent types of variable load. This simplified approach
has been adopted in CEC3.
The partial safety factors for resistance YM are given in Clause 5.1 of CEC3 and,
unlike BS 5950: Part 1 where YM = 1.0, there are three values:
Ymo 1,05 applies when plastic yielding occurs at failure
Yml 1,05 applies when either overall or local buckling occurs at failure
Ym2 1,2 applies when failure occurs on the net section of bolt holes
Separate values apply for connection design, details of which are given in Chapter 6 of
CEC3.
7.2. 7 Symbols
The symbols used to denote section properties in CEC3 are different from those used in BS
5950: Part 1 and the current structural section property tables published by SCI. A number
of 'Properties of Structural Steel Sections' tables are included in Part IV of Introduction to
Concise Eurocode 3 (CEC3)  with Worked Examples published by SCI (publication
number Pll5). Tables 7.3 and 7.4 give comparisons ofthe symbols used in CEC3 and BS
5950:Part 1.
234 Design ofStructural Steelwork
Table 7.3
Table 7.4
Note: Two new parameters hr and aLr are defined in Clauses (16) and (17) of CEC3.
The parameters combine the term ry with the buckling parameter 'U' and the torsional index
'x' used in BS 5950: Part 1. The value of aLrliLr is approximately equal to 0, 8hlt1
7.2.8 Conventions
The difference in conventions most likely to cause confusion for UK engineers is the change
in the symbols used to designate the major and minor axes of a crosssection. Traditionally,
in the UK the yy axis has represented the minor axis; in CEC3 this represents the major
axis, the minor axis being represented by the zz axis. The xx axis defines the longitudinal
axis. All three axes are shown in Figure 7 .1.
Concise Eurocode 3 CEC3 235
z
CEC3 convention BS 5950: Part 1 convention
Figure 7.1
7.2.9 Materials
The grades of steel given in Table 3.1 of CEC3 are designated as Fe 430 or 8275, both of
which represent steel equivalent to grade 43 and Fe 510 or 8355 which represent steel
equivalent to grade 50. The two designations 'Fe' and '8' represent the designations given
to types of steel in BS EN 10025:1990 and BS EN10025:1993 respectively; the 1993
version supersedes the earlier 1990 version. Both designations are given in CEC3.
The numbers following Fe and 8 are the nominal ultimate tensile strength fu and yield
strength_[y respectively, i.e.
These values must be adopted as characteristic values for design. Generally, h will be
used in calculations; however, there are circumstances, such as determining the tensile
resistance of the net area at bolt holes, whenfu will be used.
when determining the characteristic material strength, four are given in Table 6 of BS 5950:
Part 1.
Although not stated in EC3, for rolled I and H sections 't' is normally assumed to be the
flange thickness. When considering the web resistance, the value of fy corresponding to the
web thickness can be used as in BS 5950: Part 1.
236 Design of Structural Steelwork
In Clause 3.I.4 of CEC3 the value of the Modulus of Elasticity (E) is gtven as
210 Nhnm2, this differs from BS 5950: Part 1 which gives a value of205 N/mm2 •
A selection of examples are used in this chapter to illustrate the use of CEC3 for the design
of structural elements. Where necessary, extracts from appropriate Tables in CEC3 have
been included, although access to a copy of CEC3 is preferable. The clause reference
numbers relate to those used in CEC3 and not EC3 which differs slightly.
A simply supported beam ABCD supports a uniformly distributed load and two point loads
as shown in Figure 7.2. Using the design data given check the suitability of a 457 x 152 x
52 UB with respect to:
(i) shear
(ii) bending
(iii) web crushing
(iv) web crippling (this is an additional check which is not required in BS 5950:
Part 1),
(v) web buckling and
(vi) deflection
Assume the compression flange to be fully restrained.
Concise Eurocode 3 CEC3 237
W1 W1
A
~
t
Figure 7.2
D
Design Data:
Point loads: W1
Characteristic permanent load G = 4.5 kN
Characteristic variable load Q = 22.5 kN
Solution:
Extract from Table 2.1 CEC3 Concise Eurocode ... (The Steel Construction Institute)
B 13,9 kN/m C
A D
90,4kN~100,4kN
90,4 kN
xm
x = (22,8/13,9) = 1,64 m
100,4 kN
Extract from Table 3.1 CEC3 Concise Eurocode ... (The Steel Construction Institute)
I [
1+/y
:·········! I
ts40mm
......... I
~

I
flange ..___
./ ..../ ~ 
Extract from Table 5.6(a) CEC3 Concise Eurocode ... (The Steel Construction Institute)
This classification is also given in the 'Properties of Structural Steel Sections', Part IV
of SCI publication Pll5.
The Shear Area is defined in Table 5.16 and given in SCI publication Pll5 Clause
5.1 (1). For Class 1 section material partial safety factor YMo = 1, 05
(1096 X 10 3 X 275) _
Mc.Rd = 287 kNm
1,05 X 106 
where:
Sy = ft(btl tw) 0' 5 [(ytl Jywt' 5 [1 (YMoOj:Ed/ fxr) 2t' 5
for an end bearing or point load at the end of a member and
Sy = 2tt(brl fw) 0' 5 ffxrl fyw] 0' 5 [1 (YMoafEd/ fxril 0' 5
for a bearing within the length of a member or a point load within the span.
Check at support
At the end support the bending moment equals zero
a;_ M.,d _ 0
fEd  ( h f f )(h X b) 
Concise Eurocode 3 CEC3 241
where:
S, is the stiff bearing length and S/d ::; 0,2
In addition if a coexistent bending moment exists at the location where web crippling is
being checked, the following conditions must be satisfied:
At support D
Assuming the stiff bearing length 50rnm
50
== 0,12 < 0,2
407,6
Clause 5.1 (1) 1,05
_ 39,8 171,4
Fsd ~ Ra.Rd, , M.rd 5 Mc.Rd ,
Fsd
+  +
Msd
= 0,73 ~ 1,5
Ra.Rd Mc.Rd 302 287
Web is adequate under point load at point B
where:
fJA = 1,0
=
rs.
A beff X tw
beff is the effective breadth of the web taken as [h 2 + S} At the ends of a member beff
should not be taken as greater than the actual length available (~+a) ,
(where a is the distance from the centre of support to the end of the beam), nor
greater than 'h'. This is indicated in Clause 5. 7.5(2) and for various load cases in
Figure 5.17 of CEC3.
fc is the compressive strength obtained from Table 5.14 (a) or (b) using buckling curve
(c)
To obtain a value of fc from Table 5.14 requires a value of slenderness (A) which is
dependent on the conditions of lateral and rotational restraint of the flanges at the location
on the member where the force is applied.
Typical values of A, for the more common situations are given in Table 5.29.
At support D
Clause 5. 7.5(4) and Table 5.29 Assuming that the ends of the web are restrained against
both rotation and relative lateral movement.
Extract from Table 5.14(a) CEC3 Concise Eurocode ... (The Steel Construction Institute)
165 kN
Maximum design shear force Vsd = 100,4 kN < ~.Rd
Web is adequate in buckling
Since the beam has a nonstandard load case, the deflection can be estimated using the
'equivalent distributed load' technique explained in section 2.6.1 of Chapter 2, i.e.
6 ~ o,1o4 x B.MaxL2
max EJ
permanent loads:
1,0 kN 1,0 kN
A~~~~~~~~~~~~~~~~~~D
9,2 kN 6,73 kN
9,2 kN
l xm l
(9,2 + 6,4)2 (3,9 X 5,4)
B.Max = + = 26,1 kNm
2 2
Concise Eurocode 3 CEC3 245
variable loads:
5,0 kN 5,0 kN
~ D
36,4kNt=37,6kN
36,4 kN
X= (15,4/8,0) = 1,93 kN
B.Max
(36,4 + 20,4)2 + (1,92 x 15,4) = 71 ,58 kNm
2 2
Figure 4.1
L 8000
due to combined loads =    40mm
200 200
(61 + 6z) < 32 mm
Clause 5.5.5(3) indicates that lateral torsional buckling need not be checked if the spacing
of restraints is such that:
where:
k is the effective length factor from Tables 5.21 and 5.24 for beams and cantilevers
respectively.
C1 is a factor to allow for the shape of the bending moment diagram and also depends
upon the end restraint conditions. C 1 encompasses the equivalent uniform moment
factor 'm' and the slenderness correction factor 'n' used in BS 5950:Part 1. The
value of C 1 equals 1.0 for cantilevers, for other beams the values of the parameter
[k/cJ] 0•5 are given in Table 5.22 and 5.23 depending on the loading type.
hr and aLT are as defined previously and given in Section Property Tables.
f3w = 1,0 for Class 1 or Class 2 crosssections
Wez.)Wpl·y for Class 3 crosssection
We.ff)Wpl·y for Class 4 crosssections as defined in Clause 5.5.5(7)
Clause 5.5.5(4) The conditions of Clause 5.5.5(3) will be satisfied for Class 1, or Class
2 rolled I, H or channel sections provided that:
k ]0,5
[ L :::; 35,0iLr or 37,5iz for steel grade S275
c1
:::; 30,8iLT or 33,0iz for steel grade S355
ALT =
[
~I ]
ALTB
where ALTB is the basic equivalent slenderness obtained from Table 5.19 using Llhr and
aLrliLT.
In Clause 5.5.5(11) a further alternative is given to determine .fi, for a rolled section
0,5
A single span beam ABCD of 8,0 m length supports point loads at B and C as shown in
Figure 7.3. Using the design data given check the suitability of a 406 x 140 x 39 UB with
respect to lateral torsional buckling resistance.
Ar~8~~C~________,D
4,0 m 2,0 m
406 X 140 X 39 UB
Figure 7.3
Design Data:
Point load W1
Characteristic permanent load 6,0kN
Characteristic variable load 15,0 kN
Point load W2
Characteristic permanent load 6,0kN
Characteristic variable load 24,0kN
Solution:
Section properties: 457 x 152 x 52 UB
h 397,3 mm b = 141,8 mm fw = 6,3mm fJ = 8,6mm
c/tf = 8,24 dftw = 57,1 d = 359,7mm izz = 2,89mm
W.z.y = 625 cm3 Wpl·y = 718 cm3 IY.Y = 12410 cm4 Av = 27,1 cm2
aLT/iLT = 36,3 hr = 3,33 em aLT = 12,1 em
EC3 Section classification = 1. 0
30,6 kN
A~~~c~~·
4 kN 4,0 m 40,7 kN
~~~~~4~+
r1 I
Clause 5. 3 Section classification: This is given in the section property tables and is:
shown here for illustration purposes. :
Table 5. 6 Flange subject to compression :
c!t1 = 8,24 < 9,2 Flange is Class 1 :
I
Web is subject to bending 1
dltw = 57,1 < 66,6 Web is Class 1
I
1
Section is Class 1 :
A 8 c D
~ 81,4 kNm
Table 5.22 Factors [kiCJ] 0' 5 for end moment loading with k = 1,0
Table 5.22 Factors fk/C1l 0' 5 and fk/C31 End moment loading
M lf/M
(f f) k= 1,0 k = 0,85 k= 0,7
Extract from Table 5.22 CEC3 Concise Eurocode ... (The Steel Construction Institute)
0,5
[
~] ]
L = 0,957 X 4000 3828 mm, (35,0iLT) = 35,0 x 33,3 1166mm
[c kl ]o,5L > 35,0hr :. Length BC of the beam must be checked for LTB
where .tb is the bending strength and must be determined from Table 5.18(a) for rolled
sections using_fv and the modified slenderness A.Jii::
_ [k 1ct) 0 ' 5 L liLT 0,957 X ( 4000 I 33,3) 114,95

A LT [
1+ (LI aLT)
2 lo ,
25
1 + (4000 1121) 2 0 25
] ,
2,57
44,73
[
25,66 25,66
Extract from Table 5.18(a) CEC3 Concise Eurocode.,. (The Steel Construction Institute)
/b = 253 N/mm2
Clause 5.1(1) YMJ 1,05
Clause 5.5.5(10)
4000
Table 5.19 Llhr ==    120
33,3
121
= 36,3 (or from section property tables)
33,3

Extract from Table 5.19 CEC3 Concise Eurocode ... (The Steel Construction Institute)
ALTB = 109,75
Concise Eurocode 3 CEC3 251
0,5
ALT =
[
~I ]
ALTB = 0,975 X 109,75 = 104
Clause 5.5.5(11)
4000
   33,1
121
Pw = 1,0
0,5
1,0 X 0,957 X 4000 = 115
/3~5 ~I
[ ]
L/iLT 33,3
Extract from Table 5.20(a) CEC3 Concise Eurocode ... (The Steel Construction Institute)
AnNgirder supports characteristic permanent and variable loads as shown in Figure 7.4.
Using the design data provided check the suitability of the sections indicated for members
AB and CD.
Design Data:
Point load W1
Characteristic permanent load 15 kN
Characteristic variable load 30kN
252 Design ofStructural Steelwork
Point load W2
Characteristic permanent load 7,5kN
Characteristic variable load 15 kN
c D 8
6 Bays at 4,0 m
Figure 7.4
Member AB: 100 x 100 x 10 single angle, double bolted to a gusset plate
Member CD: compound strut comprising 2/150 x 90 x 12 angles double bolted
with long legs connected to both sides of a gusset plate.
Solution:
c Fco 0 8
Figure 7.5
Member AB: 100 x 100 x 12 single angle, double bolted to gusset plate
Section properties: A = 22,8 cm2
Clause 5.4.2(6) For angles connected through one leg reference is made to Clause 5.8.3.2
for bolted connections and Clause 5. 8. 3. 3 for welded connections.
2,0(e2 0,5do)tfu
singlebolted Nu.Rd =
YM2
fJ2AneJu
doublebolted Nu.Rd
YM2
fJ3AneJu
more than two bolts Nu.Rd
YM2
where:
do is the diameter of the ~olt hole
e2 is the edge distance ofthe bolt hole
/32 is the reduction factor for 2 bolts as given in Table 5.30
/33 is the reduction factor for 3 bolts as given in Table 5.30
Anet is the net crosssectional area of the angle. For an unequal angle connected by its
smaller leg, A net must be taken as the net crosssectional area of an equal of leg size
equal to the smaller leg.
Note: This differs from BS 5950:Part 1 in that it applies to all bolted angles connected by
one leg, i.e. both tension and compression members.
Clause 5.1 (1) Material partial safety factors YMo = 1,05 YM2 = 1,2
Table 5.30
Table 5.30 Reduction factors 132 and !
Pitch PI PI~ 2 5 do 2,5 <p1 < 5,0 do
2 bolts 132 = 0,4 132 = 0 1 + 1 2pl I
::?: 3 bolts /33 = 0,5 /33 = 0,3 + 0,8pl
Extract from Table 5.30 CEC3 Concise Eurocode ... (The Steel Construction Institute)
Clause 5.4.1.2(1)
Anet = [Agross (do X t)] = [2280 (22 x 12)] = 2016 mm2
Section properties:
2/150 x 90 x 12 double angles
A= 55 2 cm2
'
izz = 3,7 em iyy= 4,78 em
1/150 x 90 x 12 single angle
A= 27,6 cm2 iyy = 4,78 em izz = 2,5 em iw = 1,95 em
In CEC3 the length ofthe longer leg is denoted by 'h',and the shorter leg by 'b'.
As in BS 5950:Part 1 the thickness is denoted by 't'.
h 150
Section Classification: Table 5. 6(c)  = = 12,5 < 13,9
t 12
b+h 150+90
= = 20,0 < 21,3
12
Section is Class 3
Concise Eurocode 3 CEC3 255
Clause 5.4.3(1) Compression members
For angles connected through one leg the inplane slenderness ratios are given in Clause
5.8.3.1. These values also apply to backtohack double angles with the exception of the
check for the vv axis which is unnecessary.
Assuming the double angles are held apart by packing pieces 450 mm from each end and at
4 75 nun centres along the length.
Consider the combined double angle buckling over the full length.
4000
For the double angle: Le = 4000 mm A:w    83,7
47,8
4000
Azz =  108,1
37,0
= 475 = 19
25
f32Anetfu
Clause 5.8.3.2(2) Nu.Rd =
YM2
Design Loads:
F1 = characteristic dead load Gk= 75 kN
F 1 = characteristic imposed load Qk = 175 kN
F2 = characteristic dead load Gk= 20kN
F 2 = characteristic imposed load Qk= 75 kN
Concise Eurocode 3 CEC3 257
,/
5,0m
03 X 60 UC
Figure 7.6
Solution:
Section properties: 203 x 203 x 60 UC
h = 209,6 nun b 205,2 nun fw 9,3 nun ft = 14,2 nun
c/tf = 7,23 dffw 17,3 d = 160,8 nun izz = 5,19 nun
iyy = 9,0 em izz 5,19 em A 76,0 cm2
EC3 Section classification 1. 0
Clause 5.4.3.1(1)
7600 X 275
1,05xl03
< Nc.Rd
Section is adequate in compression
Table 5.12 Effective buckling lengths for various end restraint conditions
lyy 5000
Clause 5.4.3.2(2) ~ = =  55,6
iyy 9,0
5000
Azz = ~    96,3
lzz 51,9
Clause 5.4.3.2(5) ~ < 180 Azz < 180
Clause 6.1 0.1 (3) The effective projection should be determined from:
c  t[ /y ]0,5
3jiy MO
where:
Concise Eurocode 3 CEC3 259
k1 is the concentratiOn
. tactor whi ch.IS given
c. . by 10·  [aa1bb1 ]o,s
where 'a' and 'b' are the length and width of the baseplate respectively and 'a 1' and 'b 1 ' are
the length and width of the effective area of concrete foundation as given in Figure 6. 20 of
CEC3.
Alternatively a conservative value of~ can be taken as equal to 1,0.
Assume the characteristic strength of the grout is the same as the characteristic strength
of the concrete
/ck = C30 and /gk = 30 N/mm2
Table 6.12 /cd = fcJrc In Eurocode 2 [EC2] for concrete design the
value of Yc = 1,5
Clause 6. 10.1 (5) /cd = ~ = 20 N/mm2
1,5
Clause 6.10.1(6) /gk?:. 0,2 xfck • 0
Pi = 2/3
Clause 6.1 0.1 (5) jj = f3Jc/cd = (0,67 X 1,0 X 20) = 13,4 N/mm2
This effective projection should be considered all around each element of the column
including any baseplate stiffeners.
260 Design ofStructural Steelwork
4,2
~
59,6
Non
effective
area
290 85,9
t
59,6
~~~~~~~~~~~
4' 2 1''3:.. .::82.3',f,!":....::'...:.l''==__:__,f==..!..::._,j<,ji=..::..c.::,r
.:,
14,2 290
Figure 7.7
902 X 103
Effective bearing pressure= = 13.23 N/mm2 < jj
68,79 X 103
Adopt 290 x 290 x 15 mm thick baseplate
Clause 5. 6.1. 2
For Class 1 or 2 crosssections with low shear
Concise Eurocode 3 CEC3 261
where: MN.Rd is the reduced plastic resistance moment allowing for an axial force N.rd· The
value of MN.Rd can be determined using the approximate expression from Table
5.27. The UK NAD also permits the use of the relevant expressions given in UK
section property tables as an alternative to Table 5.27.
As with BS 5950:Part 1 both a conservative and a more exact expression is given; for
example, assuming 'low shear' for biaxial bending the Conservative Approach in Clause
5. 6.1. 2 (3) uses the expression:
The More Exact Approach in Clause 5. 6.1. 2(4) uses the expression:
(2) Buckling resistance of members with combined bending and axial tension.
Clause 5.6.2(1)
When members are susceptible to lateral torsional buckling (i.e. A, LT .fji; > 34,7 for grade
S275 steel and 30,6 for grade S355 steel), then
where:
· t l1e des1gn
1s · buckl"mg resistance
· moment = f3wh~Ly
YMJ
(3) Buckling resistance of members with combined bending and axial compression
Clause 5.6.3(1)
Three conditions are considered in Clause 5.6.3(1)
kLT = 1 1lLrNsd ~ l,Othe value of 1.1. is defined m CEC3 but can be taken
Nb.z.RdY Ml
A multistorey column in simple construction supports four beams at first floor level as
shown in Figure 7.8. Check the suitability of a 203 x 203 x 60 UC given the factored
design loads indicated.
264 Design ofStructural Steelwork
roof level
"===~====:::r,1
3,0m rM
60 kN 250 kN from
first floor level I
section xx / 8
I
35 kN
J
4,5m lW
crosssection at top of lower
slab base " A length of column AB
Figure 7.8
Solution:
Section properties: 203 x 203 x 60 UC
h = 209,6 mm b = 205,2 mm fw = 9,3 mm t1 = 14,2 mm
c/fJ = 7,23 dffw = 17,3 d = 160,8 mm izz = 5,19 mm
iyy = 9,0 em izz = 5,19 em iu = 5,53 em aLr = 64,5 em
aLrlhr = 11,7 A = 76,0 cm2
The web of the section is subject to both compression and bending and should be
classified according to the criteria given in Clause 5.3.2.
Clause 5.3.2.1 For symmetric I sections
CTw =
At section xx on column length AB
r1
1Clause 5.3.2.1
For Class 1 and Class 2 webs which are subject to bending and tension1
I
I or compress10n:
: (1) For symmetric !sections the following equation should be satisfied:
I
I
I Nsd :s; Table 5.8 value
I dtw
where:
O"w is the mean web stress, taken as +ve for compression and ve for tension
(2) For rectangular hollow sections the following equation should be satisfied:
O"w = Nsd + IMz.sdl :s; Table 5.8 value
2dtw dt(b tw)
where O"w as in (1). 1
I
I
Clause 5.3.2.2 For Class 3 webs which are subject to bending and tension or'
compress10n:
(2) For rectangular hollow sections the following equation should be satisfied:
Nsd (b fw )I Mz.sdl
+ :s; Table 5.9 value
A b VV,Lz
where Cia as m (2)
I
The section classification for Class 4 crosssections is based on the 'effective' width of:
1the compression elements and is dealt with in Clause 5.3.4. :
I I
I I
266 Design ofStructural Steelwork
Table 5.8 Limiting values of O"w (Nimm2) for Class 1 and 2 crosssections
[compression positive : tension negative]
Class 1 Class 2
dltw Fe 430 (S275) Fe 510 (S355) Fe 430 (275)
(ts 40 mm) (ts 40mm) (t < 40mm)
26
* * *
28
* 312
* p
30 273 l
* *
32 239 238
* v
34

212

207

*
 ,
   
* For Grade 275 steel with dltw = 17 3 < 26 O"w is always satisfactory
Extract from Table 5.8 CEC3 Concise Eurocode ... (fhe Steel Construction Institute)
r
Appendix F, Clauses F.4(1) and (2) of CEC3 permits the net moment at any level of a
multistorey column in simple construction to be divided between the column length above
and below that level in proportion to the stiffuess coefficients IlL of each length. This is the
same as Clause 4. 7. 7 in BS 5950: Part 1 with the exception that the approximation using
(I I L)
50% of the moment when lower s 1.5 is not included.
(I/ Ltpper
design moment about yy axis My.sd = 0,4 x 20,48 = 8,19 kNm
Concise Eurocode 3 CEC3 267
Clause 5. 6.1. 2 Resistance of Class 1 and Class 2 crosssections with low shear
A conservative check can be made using Clause 5.6.1.2(4), i.e.
If this check had failed then a more exact approach could have been used using the
expression in Clause 5.6.1.2(3).
4500
Clause 5.4.3.2(2) 50
90
268 Design ofStructural Steelwork
lzz 4500
Azz = =   = 86,7
izz 51,9
Clause 5.4.3.2(5) A, < 180 Azz < 180
N Sd + My.Sd + Mz.Sd = 705 + 1,0 X 8,19 + 1,0 X 1,05 = O,n < l,Q
Nb.min.Rd 1JMc.y.Rd 1JMc.z.Rd 1071 171,3 79,4
Clause 5.5.5(2) and (3) If ;tLT.JP: > 34,7 then lateral torsional buckling must be
checked
Table 5.22 The value of [k/CI] 0' 5 is given for end moment loading for the ratio
0,73 Assume
Concise Eurocode 3 CEC3 269
Clause 5.5.5(9)
0 73 X (4500)
' 55,3
= ~:::: = 45,7
2 0 25
[1 (450 I 645) ] •
+
25,66
~ 1,0
Section is adequate
Bibliography
5. CP 3: Chapter V Loading
Part 2: Wind Loads (1972) Superseded by BS 6399: Part 2, BSI, 1972
270
Bibliography 271
272
Index 273
Much more than documents.
Discover everything Scribd has to offer, including books and audiobooks from major publishers.
Cancel anytime.